Application Modernization

Innovation in Ireland: How Irish brands scale with Gemini Enterprise

Thu, 08 Oct 2026 12:00:00 +0000

In recent decades, Ireland has grown into a vibrant hub for global technology. As modernization accelerates, Irish organizations — spanning government agencies, established enterprise brands, and growing startups in AI and other areas — are all advancing their technical foundations. Today, leaders across the country are injecting agentic AI into every layer of their business and operational workflows.

We are seeing organizations move from early prototypes to production-ready deployments using Gemini Enterprise. This progress supports Ireland's competitive position within the European market. According to a recent Implement Consulting Group report assessing Google’s local contribution, public and private leaders estimate between 40 billion to 45 billion euros of economic potential in global AI markets.

This ambition aligns with broader government initiatives. Since assuming the presidency of the Council of the European Union for the second half of 2026, Ireland has steered higher-level technology and policy discussions, outlining a National Digital and AI Strategy to improve productivity and innovation across sectors.

To support this strategy, we are focused on making these capabilities more accessible. Our local teams are collaborating with industry leaders to implement Gemini Enterprise which offers capabilities that range from conversational AI agents and natural language data analysis, to optimizing operations with agentic ecosystems.

The economic impact is already taking shape. In 2025 alone, Google operations — including our AI and cloud services — added an estimated 10 billion euros to the Irish gross domestic product. And this momentum is shared: Organizations across Ireland are using Google Cloud to modernize infrastructure and put AI to work.

Here are several ways Irish organizations are modernizing with Google Cloud and Gemini Enterprise:

Ryanair supports millions of passengers through AI-backed operations

Ryanair, Europe’s largest airline, is modernizing its collaborative operations and infrastructure by deploying Gemini Enterprise and Google Workspace for some 35,000 employees. Flight crews, operations planners, and corporate staff will benefit from streamlined decision-making, optimized crew logistics, and enhanced corporate productivity. At the same time, a dual-cloud strategy strengthens infrastructure resilience as the airline targets growth to 300 million passengers by 2034.

Smyths Toys rethinks the customer experience with agentic support

As one the world’s largest specialty toy retailers, Smyths Toys is reshaping its customer support with Google Cloud and Gemini Enterprise. By introducing an AI agent named Codie, the retailer has resolved more than 60% of its web inquiries and reduced holiday email backlogs. Customer care teams that previously handled up to 1,000 peak-season emails a day can now focus on the roughly 50 cases that need more complex judgment.

The Irish Revenue Commissioners: Advancing public sector data

Public sector agencies are also making strides in Ireland. The Irish Revenue Commissioners — responsible for taxes, customs, and keeping Ireland’s economy running smoothly — is working with Google Cloud to migrate from legacy on-premise systems to a unified AI-lakehouse architecture designed to break down data silos and establish a single source of truth. Strict data sovereignty controls safeguard taxpayer information while Gemini Enterprise powers agentic data analysis, helping Revenue deliver secure, responsive public services.

Virgin Media Ireland: Accelerating workflows

Supported by Liberty Global's partnership with Google Cloud, Virgin Media Ireland is moving AI from exploration to production. Through modern infrastructure and automation, the team now deploys workloads three times faster, is in the process of rolling out multiple AI agents, and is launching new customer services designed to drive business growth.

Irish startups scaling with enterprise tools

While established enterprises are modernizing legacy systems, Irish startups are building on modern technology from day one. Backed by a strong local tech ecosystem, these agile teams are creating services that compete on an international stage.

Hexis: Scaling personalized sports nutrition

Hexis builds technology solutions that help athletes refine their performance and recovery. By storing and processing that real-time biometric and behavioral data in Google Cloud, Hexis uses Gemini Enterprise to power natural language conversations that explain fueling adjustments to users throughout the day. Today, the platform supports 40% of the Tour de France peloton and half of Premier League clubs during the 2026–2027 season.

Kitman Labs: Unifying performance analytics

Founded in Dublin in 2012, Kitman Labs provides sports science and performance intelligence to more than 4,000 sporting organizations across 26 countries. By combining its intelligence platform with advanced analytics and Gemini Enterprise, Kitman Labs helps teams make data-informed decisions that support athlete health, lower injury risk, and extend player career longevity.

Spryt: Automating healthcare coordination

Spryt streamlines healthcare coordination across the U.K. and U.S., using AI built on Google Cloud and Gemini Enterprise to optimize patients’ experiences, reduce missed appointments, and cut administrative burden. The company collaborates with the NHS, operates in more than 70 US clinics, and has analyzed 224 million appointment records as part of a multiyear initiative with Mayo Clinic.

IMPT: Automating carbon offsetting

By connecting AI, blockchain, and Google Cloud, IMPT embeds verified carbon offsetting into everyday hotel bookings and online purchases. Following a 15-week migration to Google Cloud, IMPT supports 8 million listings across 195 countries. With Gemini Enterprise, the company’s lean team handles about 400 daily customer chats with a six-second median response time, allowing staff to focus on growing its business rather than managing systems.

Transform ideas from planning to production

As Ireland continues to expand as a global technology hub, local organizations are demonstrating how to translate AI potential into practical results. Explore our customer stories to see how other organizations are building with Google Cloud, or reach out to our team in Ireland to accelerate your journey.

Introducing Google Cloud Modernize, transforming for (and with) AI

Mon, 05 Oct 2026 16:00:00 +0000

Today, we’re announcing Google Cloud Modernize, an end-to-end transformation portfolio to help enterprises collapse multi-year roadmaps with the power of AI.

Google Cloud Modernize brings together Google Cloud’s proven migration and modernization tools, including Migration Center, Google Cloud VMware Engine, Google Cloud Mainframe Modernization and our new EKS-to-GKE Migration Agent, into a single portfolio. With it, enterprise teams now have purpose-built agentic capabilities across infrastructure assessment, platform modernization, and application modernization.

Central to this portfolio is Modernization Hub, a newly launched in-console experience where developers and architects can analyze source code, map dependencies, and accelerate modernization for Java, .NET, and mainframe applications.

Agentic infrastructure assessment

Every transformation begins with an accurate assessment. With new Gemini capabilities in Migration Center, the Agentic Quick Estimator (GA) converts VMware inventory exports (such as RVTools) and infrastructure inputs into total cost of ownership (TCO) projections for your Compute Engine environment.

Using an interactive chat interface, teams can test real-time modeling assumptions, such as evaluating multi-region footprints or comparing BYOL licensing against pay-as-you-go, to discover contextual cost optimizations. This condenses weeks of spreadsheet modeling into a defensible business case in minutes.

Migration Center’s Quick TCO Estimator and agentic chat interface

For specialized support, we also provide a comprehensive modernization assessment at no cost through our Rapid Migration & Modernization Program (RaMP).

Platform modernization

With the rise of real-time AI agents querying backend systems, workloads increasingly require high-throughput infrastructure that removes I/O bottlenecks. Once you’ve defined your target environment using our assessments, there are several new purpose-built compute options for mission-critical workloads:

SAP S/4HANA at scale (X5 Series, GA): Delivers single-node 43 TiB memory configurations that remove the previous 29 TiB ceiling, allowing enterprise ERP estates to run without distributed partitioning overhead.
Core-optimized database performance (M4N Series, GA): Delivers 26.57 GiB RAM per vCPU paired with Hyperdisk Extreme. This prevents organizations from overprovisioning compute cores to meet memory requirements, reducing software licensing costs by more than 20% for Oracle and other core-licensed databases.
Ultra-low latency data engines (Z4D, GA and Z4M, Preview): Deliver up to 84,000 GiB and 168,000 GiB of high-speed local NVMe SSD respectively, 400 Gbps networking for both Z4D and Z4M, and RDMA support for Z4M. This throughput reduces I/O wait times and prevents query timeouts when real-time AI agents query vector stores, operational databases, and large-scale data pipelines.

Cloud elasticity for VMware environments

For organizations operating VMware estates that need the elasticity of the cloud but aren’t ready to re-architect their environment, the Google Cloud self-managed VMware solution provides administrative control on Bare Metal Z3 shapes with VMware Cloud Foundation (VCF) 9.1. Native global VPC links connect VMware estates directly to Compute Engine, Google Kubernetes Engine (GKE), BigQuery, and Gemini Enterprise, allowing teams to ground autonomous agents in operational data without code changes.

Automated container transitions to GKE

The new EKS-to-GKE Agentic Migration (Public Preview) automates transitions from AWS Elastic Kubernetes Service to GKE thanks to:

An automated pipeline: Manages discovery, Kubernetes manifest translations, storage and network mappings across clouds.
Enterprise-grade security: Built-in Human-in-the-Loop (HITL) approval gates and in-memory credential security maintain strict GitOps compliance.
A fast-track to modern runtimes: Quickly moves workloads to GKE to take advantage of low-latency model serving, autoscaling, and multi-agent orchestration.

NetEase Games demonstrated the value of this platform approach by containerizing services on GKE, reducing infrastructure scaling times from hours to five minutes during peak launches while cutting server costs by 40%:

"By integrating diverse computing instances and powerful orchestration, we have transformed our infrastructure into a competitive advantage, ensuring NetEase remains a leader in the global gaming market." - Deng Ding, Director of Site Reliability Engineering, NetEase Games

Application modernization with Modernization Hub

Landing on modern infrastructure enables teams to unlock legacy business logic and modernize their core applications. Modernization Hub centralizes several modernization tools directly inside the Google Cloud console.

.NET and Java modernization

Modernization Hub integrates the Google Cloud App Modernization CLI (CodMod), which uses Gemini to analyze large source code repositories, understand legacy application architectures,maps hidden dependencies, identifies modernization challenges and generates modernization recommendations. This enables customers to migrate legacy .NET Framework applications to modern .NET Core running on Linux containers, reducing OS licensing overhead, while also making these applications and data accessible to modern AI agent workflows.

Modernization Hub’s in-console interface

Mainframe modernization

For customers running mainframes, we provide specialized solutions to help accelerate and de-risk end-to-end application transformation to Google Cloud:

Mainframe Assessment Tool: Parses legacy mainframe codebases, extracts business rules, maps application and data dependencies to generate cloud-ready target application specifications and power agentic modernization workflows.
Dual Run: Substantially reduces cutover risk by replaying live production transaction streams simultaneously across the mainframe and the new cloud applications, verifying functional equivalence before going live.
Mainframe Connector: Copies mainframe data directly into Google Cloud services (BigQuery, AlloyDB, GCS and others), , unlocking legacy data and supports hybrid architectures.

Intesa Sanpaolo used Google Cloud mainframe modernization solutions to accelerate their core banking transformation off the mainframe and onto Google Cloud:

“To confidently move forward with Mainframe Modernization, we will need to provide confidence and assure the bank's leadership and internal control units as well as get approval from the regulators. One of the enablers for this is Google Cloud Dual Run, which is gradually providing the evidences to build such confidence to all three groups.” - Claudio Balbo, Head of IT Architecture, Intesa Sanpaolo

A proven track record and partner ecosystem

Deutsche Börse Group transitioned its mission-critical SAP S/4HANA environment and DAX® index calculations to Google Cloud, cutting disaster recovery times from hours to minutes, reducing aggregation latency by over 50%, and shortening development cycles from months to days.

To deliver these capabilities at enterprise scale, we are also collaborating with our global partner ecosystem to integrate Google Cloud Modernize with enterprise delivery frameworks, including Cognizant:

“Google’s modernization offering heralds the next frontier of AI-native modernization — enabling enterprises to autonomously decode legacy complexity and accelerate into cloud-first, intelligent architectures. Coupled with Cognizant's AI-led, governed delivery engine, we amplify this shift through agentic business processing, unlocking faster, outcome-driven value at scale.” - Nishant Upadhyaya, Global Practice Head – Digital Engineering, Cognizant

Accelerate your transformation now

Modernizing your infrastructure and applications is the baseline requirement for enterprise AI agility. Google Cloud Modernize equips teams to navigate each phase safely and efficiently as they pursue AI readiness and adoption.

Get started:

Join the webinar: Register for our upcoming session on November 17 to see live demos of Modernization Hub and our agentic migration tools.
Explore the console: Log into the Google Cloud Modernize console to access Modernization Hub.
Start planning: Visit the Google Cloud Modernize webpage, or request a modernization assessment at no cost today.

How growing UK midsize businesses are building in the AI era

Wed, 17 Jun 2026 08:00:00 +0000

The UK’s 5-million-plus small and midsize businesses and enterprises (SMBs) are the backbone of our economy. Today, we’re seeing these critical businesses begin to put AI to work, to operate more efficiently, move faster, and ultimately deliver better outcomes for their customers.

This shift is driven by tangible day-to-day results. According to recent research from Enterprise Nation published in partnership with Google, 71% of AI adopters surveyed in the UK say the technology helps them save time on routine tasks, and 64% report a direct boost in productivity. On top of this, AI-enabled productivity tools (like Google Workspace with Gemini) are delivering a 20% boost in productivity for SMBs, which effectively hands them back one full working day every single week.

At Google Cloud, we have a front row seat to this shift: SMBs have long utilized platforms like Google Workspace, and today they’re transforming with Google’s AI platform and models. In fact, we’ve seen the number of UK-based SMBs using Google Cloud AI nearly double year-over-year. This includes our Gemini models and products like Gemini Enterprise and AI Studio, which are helping SMBs do things like:

Roll out better customer support systems to help escalate and resolve customer support calls more quickly.
Automate repetitive actions in areas like payroll and accounting.
Help more employees understand and leverage data at work — even those not trained as data analysts.
Rapidly create and implement new designs for marketing collateral.
Help more people build their own AI agents to help them in their everyday jobs.
Conduct complex research projects at a speed and price point previously unavailable.

At today’s Google Cloud London Summit, we’re showcasing a number of innovative SMB customers who are actively using our AI tools to transform how they work, including companies who have recently expanded their work with us:

Neural Alpha, a sustainability fintech company, is using Gemini models to read unstructured environmental and corporate sustainability reports to automatically find and organize thousands of key facts, cutting months of slow, manual research down to a fraction of the time.
Sep 2, a digital security provider, uses Gemini Enterprise to deploy autonomous AI agents for 24/7 threat monitoring — accelerating incident detection and quickly neutralizing security threats for its customers.
Sunhouse, a strategic brand design agency, uses Gemini Enterprise to easily find archived design work stored on Google Drive, enabling its teams to spend less time hunting for files and more time growing its business with global brands.
Terrapinn, a global B2B events company, is transforming its operations by leveraging Gemini models, NotebookLM, Looker, and BigQuery to turn manual tasks into automated workflows, accelerating how its teams design, market, and deliver world-class conferences.
VoCoVo, a telecommunications provider, is integrating Google Cloud AI across its systems to turn isolated data into actionable intelligence and build autonomous workflows, streamlining routine operations so their team can focus on high-impact innovation.

Empowering Your Team: AI Upskilling Resources for Growing British Businesses

To help midsize teams maximize their impact and confidently navigate the modern AI landscape, we’ve developed a suite of dedicated, no-cost upskilling resources. Whether you want to train your existing teams or democratize data tools across your entire workforce, these programs will help you build an AI-ready organization:

SMB-Focused Programs: Explore our new SMB Learning Path or enroll in the Gemini Enterprise Agent Ready (GEAR) program for specialized training in agentic AI.
Google Skills for Organizations: Access our no-cost, on-demand learning platform featuring over 3,000 AI courses and hands-on labs created by experts at Google Cloud and Google DeepMind.
Get Certified: Ready to validate your team's expertise? This premium, cohort-based program offers instructor-led training, technical mentorship, and AI-infused skill badges designed to prepare your team for industry-recognized certifications.

By offering a full suite of SMB technology and training — from productivity in Workspace, to all our Ads services, and now powerful AI tools — Google is helping small and midsize firms thrive, no matter where the future takes us.

Cool stuff Google Cloud customers built, May edition: Agentic algorithms for supply chains; virtual try-on APIs; robotic camera operators & more

Fri, 29 May 2026 16:00:00 +0000

AI and cloud technology are reshaping every corner of every industry around the world. Without our customers, who are building the future on our platform, there would be no Google

Cloud. In this regular round-up, we dive into some of the exciting projects redefining businesses, shaping industries, and creating new categories.

For our latest edition, we learn how Urban Outfitters sped up its order management; BASF uses AlphaEvolve algorithms to map global supply chains; the unification strategy for UKG’s workforce intelligence; WPP’s secrets to training humanoid robot camera operators; how Breuninger piloted Virtual Try-On APIs; creating automated video clips with Glance; and Movix improves the production of dental aligners.

Be sure to check back next month to see how more industry leaders and exciting startups are putting Google Cloud technologies to use. And if you haven’t already, please peruse our list of 1,302 real-world gen AI use cases from our customers.

Urban Outfitters saves big by migrating order management

Who: Urban Outfitters, Inc. (URBN), the popular clothing and home goods retailer, relies on IBM Sterling OMS as the nerve center of its global ecommerce operations. However, the foundation of this critical system — a massive 11TB Oracle database — was increasingly becoming a bottleneck.

What they did: URBN completed a major infrastructure upgrade, migrating its IBM Sterling OMS from an Oracle database to Google Cloud's AlloyDB for PostgreSQL. To enhance performance and provide high availability and scalability, the AlloyDB deployment architecture includes two read replicas, providing low-latency access to data for reporting and analytics. Google Cloud and IBM teams also assisted URBN in a rigorous, iterative switchover testing strategy.

Why it matters: The migration to AlloyDB has fundamentally reshaped URBN’s data strategy, delivering a more favorable total cost of ownership through an optimized storage and compute architecture, without sacrificing performance or reliability. Furthermore, the shift to a PostgreSQL-compatible database gave URBN the flexibility of an open-source ecosystem, providing freedom from vendor lock-in, as well as significant speed improvements that enhanced responsiveness.

Learn from us: "URBN’s successful migration serves as a blueprint for organizations looking to modernize their mission-critical infrastructure and future-proof their environment for AI expansion. This journey proves that even the most complex, mission-critical migrations can be achieved through deep cross-organizational partnership and a phased, risk-mitigated approach." – Rob Frieman, CIO, Urban Outfitters & Raj Pai, VP, Product Management, Databases, Google Cloud

BASF manages supply chain decisions with AlphaEvolve

Who: BASF Agricultural Solutions manages a complex network of 180 production sites with more than 5,000 distinct value chains. Currently, human planners make thousands of local decisions every day on what to produce, when to produce it, and how much safety stock to hold.

What they did: To understand how local decisions ripple across their entire global network, BASF turned to AlphaEvolve on Google Cloud to build a digital twin of their supply chain. In collaboration with Google Cloud and prognostica GmbH, BASF fed the model three years of historical data and then generated variations of the code, mutating the logic to see if it could simulate a supply chain that matched the real-world historical data.

Why it matters: By running thousands of experiments, AlphaEvolve developed a clear, human-readable algorithm that explains how the BASF network truly operates. The final algorithm successfully mirrored the actual historical performance of the supply chain, significantly reducing the error rates compared to the initial seed model. It automatically discovered factually correct, domain-specific supply chain rules, providing a clear foundation for optimizing asset utilization globally.

Learn from us: “We had several attempts to build a digital twin. … By using AlphaEvolve, we cannot only map the complex network based on system data, but at the same time understand and copy the human decisions that drive our daily operations.” – Dr. Goetz Krabbe, vice president for global supply chain at BASF

UKG unlocks real-time workforce intelligence at scale

Who: UKG is one of the leading providers of human capital management (HCM) and workforce management (WFM) solutions, but years of growth led to backend sprawl. They have 126 application teams, dozens of tech stacks, and more than 12,000 database instances.

What they did: To bring the full UKG suite onto one real-time foundation, the company built People Fabric, a new data and intelligence platform powered by AlloyDB for PostgreSQL and the just-announced Agentic Data Cloud. They created a custom change data capture (CDC) framework to extract changes from existing operational databases, and for larger analytical workloads, the same data flows into BigQuery, while Cloud SQL holds the metadata and tenancy context.

Why it matters: People Fabric gives UKG a complete and consistent view of people, work, pay, and culture data that’s updated continuously and ready for AI to use in real time. For engineering teams, People Fabric acts as a database-as-a-service that accelerates development and supports modernization without customer disruption. Additionally, migrating core person and employment data off their on-prem monolith has generated cost savings significant enough to fund half of People Fabric.

Learn from us: “As we continue expanding People Fabric, we’re laying the groundwork for deeper agentic automation, more responsive analytics, and a growing set of AI-driven capabilities — all on a trusted, scalable foundation built for what’s next.” – Radhi Chagarlamudi, Group Vice President, Product Engineering, UKG & Heather White, Cloud Data Architect, Google Cloud

WPP accelerates humanoid robot training 10x with G4 VMs

Who: WPP is one of the world’s largest marketing organizations, handling $70 billion of media for enterprise clients. They work on some of the most complex commercial film shoots and were eager to test the viability of robotic cameras to capture more footage, but this required complex training of physical models AI.

What they did: WPP used the new G4 VM instance powered by NVIDIA RTX PRO 6000 Blackwell on Google Cloud to tackle the unique challenges of training physical AI for robotics in videography settings. After capturing human motion with the OptiTrack mocap system, they undertook reinforcement learning using the AI Hypercomputer together with the NVIDIA Isaac Sim image. MuJoCo, an open source physics engine by Google DeepMind, was a critical piece of simulation software that validated accuracy continuously, in real-time.

Why it matters: WPP was able to utilize a P2P topology that moves data directly between GPUs without the bottleneck of central processing. They saw speed increases in excess of 10x, taking training times down to less than one hour. Through high-volume simulation, the humanoid robots learned how to respond to small changes and bridge the tough "sim-to-real" gap, helping ensure the robot's simulated adaptability translated to safety and stability in the real world.

Learn from us: "Our process for mastering complex, natural movement on a film set can be replicated across industries to overcome the massive computational complexity of training robots." – Perry Nightingale, SVP of Creative AI, WPP

Breuninger boosted sales with its "be your own model" AI

Who: Breuninger, a fashion and lifestyle company based in Germany, thought emerging generative media models could be a good fit to answer the question every online fashion shopper asks: "How will this look on me?"

What they did: Working with Google Cloud, they built a virtual try-on experience that lets shoppers see high-end fashion on their own bodies using a simple selfie. Using the Virtual Try-On (VTO) API, Breuninger’s data team worked directly with Google’s engineers to test and refine the technology in three stages, ultimately moving from pre-selected models to a user-first, selfie-based approach. The project was also part of Breuninger’s move to a Flutter-based platform, which helped the team move from its vision to a live launch in only three months.

Why it matters: During a six-week A/B test over Black Week and the holiday season, the team found that shoppers who used the virtual try-on converted purchases at a higher rate than those who didn't. Customer surveys reinforced the numbers: shoppers responded well to the high image quality and the personalized experience.

Learn from us: “Breuninger continues to refine the experience based on how customers actually use virtual try-on in everyday shopping — the same user-first approach that shaped the project from the start.” – Daniel Rascher, Senior Product Owner, Breuninger & Dr. Michael Menzel, Customer AI Specialist, Google Cloud

Glance turns hours of video into mobile-ready clips

Who: Glance, a mobile-first content platform, processes 1-2 hour videos from sources like podcasts, news reports, movies, and web series, and transforms them into 30 to 180-second vertical clips optimized for mobile lock screens.

What they did: The goal was to create a complete pipeline that takes a long-form landscape video (16:9) and outputs multiple ready-to-publish short-form portrait videos (9:16). The final technical solution uses Google Cloud Speech-to-Text v2, Gemini, and the Google Vision API, combined with custom video manipulation using Samurai (an open-source object tracking tool), OpenCV and MoviePy. The process involves audio extraction, speech-to-text transcription, and using Gemini 2.5 Flash to analyze transcript text and identify optimal start and end timestamps for short video clips.

Why it matters: With daily volume projected to grow from 3,500 to over 10,000 videos per day, manual editing wasn’t a realistic path forward. Glance’s video pipeline demonstrates what becomes possible when AI handles the repetitive, judgement-intensive work of video editing. The system transforms thousands of long-form videos into mobile-ready clips each day, preserving narrative context while optimizing for vertical viewing. Rather than choosing between scale and quality, automated pipelines can deliver both.Learn from us: “Glance’s video pipeline demonstrates what becomes possible when AI handles the repetitive, judgement-intensive work of video editing. … The approach offers a template for any organization sitting on long-form video archives. Rather than choosing between scale and quality, automated pipelines can deliver both.” – Himanshu Aggarwal,

Machine Learning Engineer, Glance & Sharmila Devi, AI Consulting Lead, Google Cloud

Movix fills a gap in dental skills with specialized agentic AI

Who: Movix is building one of the first agentic AI solutions for dental appliance manufacturers and dental labs, to help solve a serious shortage of skilled dental technicians in aligner manufacturing.

What they did: Movix developed custom models for deep learning, computer vision, and 3D mesh analysis over a five-month period, using Google Cloud infrastructure. Once defects are detected, they use the Gemini Enterprise Agent Platform to generate client-facing feedback that reads as if it came directly from a human technician. Their 3D models use Cloud Run with L4 GPUs for the massive compute power required, and they use Compute Engine VMs to run experiments and train models.

Why it matters: Movix’s agentic solutions automate data entry and quality control, which are traditionally manual, time-consuming, and error-prone tasks. The automation and higher level of accuracy the QC agent delivers can save $300 per remake for an aligner manufacturer, and speed up the appliance manufacturing process with quicker turnaround times.

Learn from us: “We plan to build hybrid solutions … designing an architecture that connects our cloud-based AI agents with older, on-premises software that many conservative labs still use — through lightweight local connectors and standardized APIs. This will allow us to access a large market segment that has not yet migrated to the cloud.” – Marina Domracheva, CEO, Movix & Bakit Dzhumagulov, CTO, Movix

Google Cloud and SAP unveil blueprint for the Agentic Enterprise

Wed, 22 Apr 2026 12:00:00 +0000

Google Cloud and SAP are deepening their partnership, working together to embed Gemini AI directly into the core business processes of the world's largest enterprises.

A new foundation for AI: Today at Google Cloud Next '26, we are joining SAP to showcase the latest breakthroughs for the Agentic Enterprise from our ongoing partnership.

These include the Unified Data Foundation, which integrates SAP and non-SAP data into a bi-directional source of truth, transforming routine cloud modernization into a strategic engine for data-driven value. Our new zero-copy data sharing capabilities further simplify architectures by removing the friction of data movement, ensuring the high-fidelity data access and real-time reliability essential for building mission-critical AI workloads.

We’re also supporting new capabilities in the SAP Engagement Cloud to deliver the latest in AI creation, ideation, and learning tools. These platforms are available not only to SAP employees but also as an offering for SAP’s own customers. This suite of AI tools will help drive agentic understanding and adoption across the enterprise.

Why it matters: This shift closes the gap between data storage and AI innovation. It empowers organizations to move from simply managing infrastructure and start deploying intelligent agents that can autonomously execute complex, multi-step tasks.

The Unified Data Foundation in particular offers not only valuable insights for existing workflows but also helps power the outputs of AI tools and agentic workflows.

The big picture: The deepening partnership is already transforming how SAP delivers value to its own customers.

Driven by an internal mandate to maintain market relevance and build a competitive advantage, SAP's Engagement Cloud division partnered with Google Cloud to build next-generation agentic solutions. These include agents for dynamic content development, generation of marketing briefs and visual concepts, and collaborative multi-agent execution.

What they're saying: “Agentic AI only creates value when it’s grounded in trusted data and directly connected to execution, which is what makes our partnership with Google Cloud so exciting for businesses. By embedding generative and agentic intelligence directly in Engagement Cloud, Google and SAP are giving marketing teams a practical way to turn real-time intelligence into meaningful customer interactions.” – Joanna Milliken, Head of Engagement Cloud, SAP

What's new:

The zero-copy revolution: SAP BDC Connect for BigQuery delivers bidirectional, zero-copy data sharing. You can unify your entire data footprint without moving or duplicating massive datasets.
Cortex Framework: New solution accelerators embed rich metadata directly into BigQuery. This grounds Gemini agents in an accurate enterprise context, heavily mitigating AI hallucination risks.
Elevated 99.95% SLA: An industry-leading uptime guarantee for RISE with SAP on Google Cloud, driven by AI-powered predictive outage prevention.
SAP Sovereign on Google Cloud: Customers with strict data residency mandates can now run SAP's S/4HANA Private Cloud Edition directly on Google's sovereign infrastructure.

By the numbers: Customers can achieve up to a 54% lower total cost of ownership (TCO) for data analytics. Hosting SAP workloads on Google's energy-efficient public cloud also instantly accelerates corporate ESG targets.

What's next: SAP BDC Connect for BigQuery will soon be available in select data centers, with broader availability in the second half of 2026.

Go deeper:

Building the Agentic Enterprise with Google Cloud partners and a $750M innovation fund

Wed, 22 Apr 2026 12:00:00 +0000

We are now seeing the Agentic Enterprise become reality for customers, and this week at Next ‘26 we are announcing exciting, new innovations to help customers accelerate agentic AI even further.

Our partners play a critical role in enabling the Agentic Enterprise, and today we are also announcing new resources, technologies, and deep technical partnerships to ensure we offer customers the industry’s most capable partner ecosystem for the agentic era, including:

A $750 million partner fund for agentic development applicable across global consulting firms, software partners, and our channel partners.
New ways for customers to deploy partner agents in Gemini Enterprise.
Deeper and more technical partnerships with global consulting firms to support customers, including with new teams of Google forward deployed engineers.
Integrating Gemini models more deeply into enterprise platforms from Palantir, Salesforce, SAP, ServiceNow, and more.
More AI-powered features in Google Cloud Partner Network to help our partners deliver high quality services.

Investing to accelerate AI agent development

We are committed to offering customers the most AI-capable partner ecosystem in the industry. To empower our partners to drive real transformation in the agentic AI era, we are launching a $750 million innovation fund to accelerate agent development and deployment globally, applicable to every business process, function, and industry.

This funding will support a wide range of activities including:

Hands-on support for software companies to build AI agents into their products with the Gemini Enterprise Agent Platform and bring them to market through our Agent Marketplace and through the new Agent Gallery in Gemini Enterprise.
Expert Google forward deployed engineers (FDEs) who will partner with major systems integrators to help their customers solve deep technical challenges and deploy Google AI more rapidly.
Deployment and usage incentives to help services partners thrive in the agentic era.
Training, technical development initiatives, and workshops to help partners build and deploy agents for customers using Gemini Enterprise Agent Platform.

Surfacing partner-built agents in Gemini Enterprise

At Next, we’re announcing Gemini Enterprise Agent Platform, a comprehensive platform to build, scale, govern, and optimize agents. It includes Agent Gallery, where customers can browse a highly vetted set of agents built by our partners.

Today, Agent Gallery provides access to a wide range of third-party agents. These agents have been built on top of our secure, enterprise-grade infrastructure, meaning customers can deploy them within their businesses with the highest levels of governance and confidence. Today, this includes agents built by Accenture, Adobe, Atlassian, Deloitte, Lovable, Oracle, Palo Alto Networks, Replit, S&P Global, Salesforce, ServiceNow, Workday, and more.

Empowering global consulting partners to drive AI transformations

Today, Google Cloud’s global consulting and systems integrator partners offer customers more than 330,000 experts trained on implementing Google AI. At Next, we are expanding our partnerships with every major systems integrator, including:

Accenture is helping enterprises drive AI-powered reinvention and business value faster and at scale with the launch of a first-of-its-kind Gemini Enterprise Acceleration Program. The program brings elite engineering and forward deployed engineers from Google Cloud and Accenture directly to customers.
BCG is expanding its partnership with Google Cloud to accelerate Gemini Enterprise transformation, helping organizations deliver at-scale agentic adoption.
Capgemini is establishing a Google Cloud AI Enterprise Hub to accelerate enterprise-scale adoption of Gemini Enterprise.
Cognizant is launching a dedicated Gemini Enterprise practice group to accelerate enterprise adoption of Gemini Enterprise.
Deloitte is forming a dedicated Google Cloud Agentic Transformation practice focused on Gemini Enterprise and will roll out Gemini Enterprise to more than 100,000 of its own teams.
HCLTech is launching a Gemini Enterprise Business Unit to accelerate the development and adoption of industry-specific solutions built on Gemini Enterprise.
Infosys is leveraging Gemini Enterprise within its Infosys Topaz platform and is equipping more than 100,000 Infosys developers across Infosys’ global delivery teams with Gemini Enterprise.
KPMG is deploying Gemini Enterprise with a life sciences company and launching a new Financial Close Companion agent built with Workday and Google Cloud.
Kyndryl is deepening its Google Cloud partnership with expanded Google Distributed Cloud services for sovereign, AI-ready applications.
McKinsey is launching the McKinsey Google Transformation Group to accelerate enterprise AI outcomes with Gemini Enterprise, combining its strategic expertise with Google's AI stack to help organizations scale agentic transformation
PwC is launching a dedicated Google Cloud AI Center of Excellence to help organizations scale AI adoption, pairing industry expertise with Gemini Enterprise to deploy AI agents that reason, act, and automate processes at scale.
TCS is launching new agentic AI offerings and a dedicated Gemini Enterprise practice, featuring more than 3,000 industry-focused AI agents and an expanded global network of Gemini Experience Centres to accelerate AI-native, autonomous enterprise operations.

Additionally, we will begin to embed teams of Google Cloud engineers with a subset of global partners, including Accenture, Capgemini, Cognizant, Deloitte, , HCLTech, PwC, and TCS in order to help their customers more rapidly prototype and deploy AI agents within their businesses. AI-native services partners, including Altimetrik, Artefact, Covasant, deepsense.ai, Distyl.ai, Northslope, Quantium, Tribe.ai, and Tryolabs will launch Gemini Enterprise practices, receiving credits for sandbox development, technical upskilling, and referral opportunities.

We are also rolling out a new program offering early model access for a select group of partners, including Accenture, Bain & Company, BCG, Deloitte, and McKinsey, who will be able to preview and begin building with pre-release versions of upcoming Google DeepMind models.

Bringing Gemini to More Customers through Popular SaaS Platforms

Already, many of the world’s leading agentic SaaS and AI platform companies integrate Gemini into their products. At Next, we’re expanding these integrations even further, including:

Atlassian is bringing Gemini 3 Flash to Rovo and integrating multimodal capabilities into Remix in Confluence, helping teams instantly transform text-based documentation into high-fidelity diagrams and charts for faster stakeholder decision-making.
Box is launching new Box Agents powered by Gemini 3 Flash and Gemini Enterprise, helping enterprises transform static files into actionable intelligence by natively integrating AI orchestration into their secure content management workflows.
DocuSign is using Gemini to power new features that summarize complex agreements, identify key clauses, and help users understand the implications of their contracts.
Oracle is launching the Oracle AI Database Agent for Gemini Enterprise. This new agent enables end users to ask business questions of their Oracle data in natural language in Gemini Enterprise, without needing to write SQL or understand the underlying data model.
Palantir is adding Gemini and BigQuery integrations for commercial customers, enabling customers to connect best-in-class models to their most critical AI workflows and operations.
Salesforce is adding native Gemini support to its Atlas Reasoning Engine. This enables Agentforce to “see” across text, image, and video formats, drawing from years of customer history to accurately solve complex problems. This means faster, smarter resolutions, building upon the success thousands of customers are already seeing from Gemini within Agentforce to build prompts.
SAP is integrating Gemini Enterprise into its Engagement Cloud to deliver AI-powered customer service and sales insights alongside creative tools for image and text generation.
ServiceNow is integrating its AI agents with Gemini Enterprise, bringing autonomous operations to the world’s largest enterprises

Building a partner channel for the agentic era

Our new partner program, Google Cloud Partner Network, is designed to help partners thrive in the agentic era. Last year, we used AI to unlock deep insights across our partner tools; now, we are building the agentic workflows that turn those insights into autonomous growth. Key updates include:

The Partner Agent: Integrated into the Partner Network Hub, this agentic tool acts as a central orchestrator for the partner experience. Beyond answering questions, it actively guides partners on next steps, summarizes complex assets, and provides real-time coaching for registrations and statements of work.
The Agentic Earnings Hub: Here, partners can find new capabilities to auto-draft statements of work and monitor consumption milestones to auto-generate claim requests. When paired with the Earnings Potential Modeler, these tools provide contextual recommendations to map every available incentive down to the individual client level.
Partner Finder: We are also extending this intelligence to customers, turning discovery into a conversational experience where natural language prompts pinpoint the ideal partners for the most hyper-specific workloads.

Finally, we’re honored to highlight the winners of Google Cloud’s 2026 Partner Awards, which celebrate the transformative impact and incredible value our partners have delivered for customers over the past year. Our ecosystem continues to evolve to meet the needs of businesses across every industry, and we are constantly impressed by their ability to solve complex, global challenges using our technology. To discover more about these exceptional achievements, please read our full list of partner award winners.

I can’t wait to meet with thousands of you this week to build the future of the Agentic Enterprise!

Building the agentic future: A spotlight on Google Cloud’s media & entertainment partner ecosystem

Thu, 16 Apr 2026 13:15:00 +0000

As we gather in Las Vegas for NAB Show 2026, the industry conversation has shifted. We are no longer asking if AI works; we’re now focused on how it scales. The era of AI experimentation is over — production-grade execution is here.

At Google Cloud, we believe no studio or broadcaster should have to build this future in isolation. Our mission is to provide the agentic platform and AI and cloud tools that allow our partners to innovate at the speed of ideas — from the tools used in the edit suite, to the technology that delivers video to millions of viewers worldwide.

Enhancing production: From manual tasks to intelligent assistants

Modern creative workflows are often slowed down by manual technical tasks. Google Cloud is working with ecosystem leaders to integrate advanced AI capabilities directly into the core of production software, so creators can focus on their artistry, not tedious tasks. With AI acting as a proactive assistant within the creative suite, production teams can significantly reduce the time between a raw idea and a finished frame.

Avid: With the launch of Content Core on Google Cloud, Avid is delivering a truly cloud-native studio. And by integrating multimodal AI search into Media Composer, editors can find the exact frame they need using natural language, turning hours of logging into seconds of discovery.
Backlight: Backlight makes complex media workflows simple for teams of all sizes, from production through monetization. Built on Google Cloud with the Video Intelligence API, Backlight's Iconik platform automatically adds searchable metadata upon upload. Customers see up to 50% shorter production cycles and save up to 60% on storage by deeply understanding their media libraries, reducing duplications, and optimizing asset placement.
Brahma.ai: Brahma AI, an enterprise AI content platform, is powering high-fidelity digital likenesses across retail, entertainment, and healthcare, making them interactive and intelligence-driven within a secure and governed framework.

Unlocking content value: From static archives to active assets

Data is only as valuable as the insights you can extract from it. Our partners, listed on the Google Cloud Marketplace, are using generative media models to transform massive, static archives into searchable, revenue-generating engines. By making every frame discoverable, we’re helping media companies turn decades of history into immediate opportunities.

Ateme: Ateme helps simplify global distribution with its new generative AI-powered subtitling solution, which can significantly reduce the manual labor of localizing different media types.
Perfect Memory: Perfect Memory helps customers turn traditional storage into a context-aware knowledge engine. The platform understands the relationships between athletes, historical events, and emotional nuances — transforming massive media archives into an intelligent library that lets creative teams instantly surface the perfect content for any story.
VionLabs: Working with companies like Cineverse, Plex, and Crunchyroll, Vionlabs uses AI to analyze and index content libraries — making video assets more accessible and enabling metadata generation. By understanding the specific mood and aesthetic of each scene, Vionlabs helps streaming platforms move beyond basic genre tags toward more nuanced, sentiment-driven content discovery and marketing.

Scaling global reach: From simple streams to audience growth

To succeed today, media companies must provide a smooth viewing experience and easy payment options. Our ecosystem provides the tools to grow a company’s reach and maximize the value of every subscriber through reliable, high-quality delivery.

Bending Spoons: By leveraging the global scale of Google Cloud, Bending Spoons’ properties such as Brightcove and Vimeo are delivering professional-grade tools for large enterprises, SMBs, the next generation of creators, and more. Its platforms ensure that high-quality video production and distribution are accessible to everyone, from global brands to independent storytellers.
Bitmovin: Bitmovin enables streaming services to scale efficiently while delivering a premium experience across the widest range of devices. By combining real-time observability with AI-driven insights, media teams can proactively optimize engagement and monetization. Furthermore, Bitmovin’s advanced encoding ensures superior visual quality at lower bitrates, supporting everything from high-demand Video on Demand (VOD) to massive, 24/7 live events.
Evergent: Evergent automates complex billing and monetization workflows for AI-powered revenue management. Media and telecommunications companies can use Evergent’s tools to maximize subscription growth and improve long-term customer retention through personalized and agile payment offers.
Harmonic: Harmonic is helping major broadcasters like Grupo Globo modernize their operations. By integrating new digital broadcast capabilities into their cloud-based streaming solutions, Harmonic provides leaders with a faster, more efficient path to manage video processing and delivery at a massive scale.

Ensuring reliability: From infrastructure to a foundation of trust

High-quality content requires a high-performance foundation. We are partnering with infrastructure leaders to ensure that even the most complex global broadcasts remain stable, secure, and responsive.

Zixi provides the broadcast-grade transport and workflow automation needed to move professional video across any network. By offering centralized control and complete visibility into the delivery process, Zixi ensures that leaders like Fubo can manage high-stakes, broadcast-quality live events without the risk of a signal drop.

Visit the ecosystem in action

The strength of our ecosystem is its integration across all aspects of the media and entertainment landscape. From the cameras, to the cloud, to the viewers' screens, these partners represent the future of a more creative, efficient, and agentic media industry.

Visit the Google Cloud Booth (West Hall, #W2731) at NAB Show from April 19-22 to see many of these partners in action through live demonstrations and theater sessions.

aside_block: <ListValue: [StructValue([('title', '2026 AI Agent Trends in Media and Entertainment'), ('body', <wagtail.rich_text.RichText object at 0x7fe0b8223b10>), ('btn_text', 'Read it now.'), ('href', 'https://cloud.google.com/resources/content/ai-agent-trends-media-entertainment-2026'), ('image', <GAEImage: Confirmation email_500x450>)])]>

Cool stuff Google Cloud customers built, April edition: BMW big on SLMs, MLB’s Scout Insights AI, personalized resort experiences

Wed, 15 Apr 2026 16:00:00 +0000

AI and cloud technology are reshaping every corner of every industry around the world. Without our customers, who are building the future on our platform, there would be no Google Cloud. In this regular round-up, we dive into some of the exciting projects redefining businesses, shaping industries, and creating new categories.

For our latest edition, we learn why BMW Group is experimenting with small language models (SLMs); catch AI-powered commentary from Major League Baseball; hit the slopes with Vail Resort’s AI concierge; build an intelligent grid with CTC Global; witness how ID.me created secure global scale; and see how Manhattan Associates supply chain tools now handle 1 billion daily API calls.

Be sure to check back next month to see how more industry leaders and exciting startups are putting Google Cloud technologies to use. And if you haven’t already, please peruse our list of 1,001 real-world gen AI use cases from our customers.

BMW tests the big potential of small models

Who: As one of the world’s leading providers of premium cars and motorcycles, BMW Group is always at the forefront of automotive technology. This ethos pushed the company to test what type of AI language models are ideally suited to driving situations, where access to cloud-based LLMs isn’t always possible.

What they did: BMW Group wanted to explore the potential of small language models (SLMs), which could run within the limited hardware on a vehicle. Finding the right trade-off between size and capability requires careful optimization, and the sheer volume of viable combinations renders manual searches for the optimal configuration an incredibly tedious, if not impossible, undertaking. To overcome this challenge, BMW and Google built automated, reproducible workflows through executable pipelines using Vertex AI.

Why it matters: The path from a general-purpose LLM to a specialized SLM isn’t straightforward. Every choice — from type of quantization to characteristics and contents of the fine-tuning domain-specific dataset — affects the quality and efficiency of the final model. This creates an exponential range of configurations, each with different trade-offs. It’s a great example of using AI to scale an optimization problem for other AI.

Learn from us: “With automated pipelines, we can rapidly adapt models to our domain and rigorously test and evaluate them against domain-specific benchmarks. This allows us to iterate and optimize models in hours rather than days.” – Dr. Céline Laurent-Winter, vice president, Connected Vehicle Platforms at BMW Group

MLB Scout Insights: AI-powered color commentary

Who: Major League Baseball is famous for its colorful announcers. Now, MLB is bringing more baseball color straight to your pocket, and Gemini is helping give it a voice.

What they did: Each season, millions of baseball fans use the MLB app and tap over to the Gameday feature for live, up-to-the-pitch action across more than a dozen games. Starting this season, the league launched MLB Scout Insights in Gameday, which uses Gemini models to quickly scan decades of game and player data, cross-references it with situational game scenarios, and then delivers game-relevant context during key matchups.

Why it matters: Given the sport’s storied history, 162-game regular season, and global reach, baseball fans are among the most sophisticated and passionate out there. To keep them engaged with Gameday and the MLB app, the league wanted to deliver insights that truly felt meaningful and interesting. Building the tool meant answering a rather squishy question: What makes an insight actually insightful, not just an accurate fact, and how can an AI learn that distinction? The answer came from some clever “surprisal” analysis.

Learn from us: "With Scout Insights, every fan can feel like the smartest person in the stands, at the water cooler, or on the couch. It’s about deepening connections to the game, and sharing that passion with others. That’s the magic of sports, and we’re making more of it possible with the magic of AI." – Josh Frost, senior vice president of product & Matt Graser, director of engineering, Major League Baseball

Vail Resorts makes personalized AI assistance easy

Who: Vail Resorts operates some of the most iconic and beloved mountain destinations in the world, including Whistler Blackcomb, Park City Mountain, Stowe, and Crested Butte.

What they did: Vail Resorts launched My Epic Assistant during the 2024-2025 snow season, and expanded it this year to add even more AI-powered chat features powered by Google’s powerful Gemini models. The result is an agentic guide to the slopes that can help skiers and snowboarders decide on the right season pass, share the latest snow report, check on lesson preparations, or suggest a good stop for cocoa.

Why it matters: Vail Resorts wanted more than a chatbot; they wanted a digital concierge that understands the nuance between a powder day at Whistler and a family trip to Beaver Creek. As the company implemented and refined personalization, improved search, summary capabilities, and conversational flow within My Epic Assistant, the app has delivered a 45% reduction in escalation to human agents since launch.

Learn from us: "Utilizing tooling from Google Cloud, we could lean into agentic design patterns that gave us a way to unlock natural, personalized conversations. These boosted customer satisfaction, while reducing the need for manual intent design. These tools also let us combine flexibility and control to enable the assistant to respond fluidly but always within the boundaries of our brand, policies, and product strategy.”— The Vail Resorts technical team

CTC Global turns the smart grid into an intelligent one

Who: CTC Global is a leading manufacturer of advanced transmission conductors and power lines. While many nodes in the grid contain IoT sensors, it recognized a literal gap in the transmission lines themselves.

What they did: CTC’s new GridVista platform threads fiber-optic cable into its high-strength carbon fiber composite core, and connects these to monitoring technology built with AI and monitoring technology from Google Cloud and Tapestry. With GridVista, CTC can turn every inch of transmission into a smart sensor.

Why it matters: GridVista gives CTC grid operators an accurate and reliable view of what’s happening across the entire line — based on actual, real-time data from the entire length of the conductor, not point estimates from a static model of line conditions or the occasional clamped-on sensor. This means they can significantly improve safety, manage costs, increase the line’s capacity to transmit power, and enhance reliability with more precise insights about events that might trigger an outage.

Learn from us: “This awareness allows for a grid that can truly sense its own health in real time and provide unprecedented awareness of conditions on the entire line. Whether that’s real time storm impacts, ice load, wind load, branches on the wire, or temperatures on or under the line. The GridVista system truly represents next generation capabilities. ” — J.D. Sitton, CEO, CTC Global

ID.me reduces risk while scaling past 160 million users

Who: ID.me is transforming digital identity security for the modern era, offering a single login that lets you easily prove you’re you across a wide range of platforms and wallets.

What they did: ID.me currently serves more than 160 million users, including as many as 40,000 at any time, so they can prove their identity online as easily as flashing their driver’s license in person. Over the last two years, ID.me migrated more than 50 terabytes of data across 15 database instances to Google Cloud with minimal downtime. They also introduced a two-tier architecture with Cloud SQL supporting its smaller and more standard services, while AlloyDB runs heavier workflows that form the backbone of the ID.me platform.

Why it matters: AlloyDB AI has allowed ID.me to scale its systems to handle 10X-20X of what was possible before — and at a lower price to boot. That responsiveness and reliability led the U.S. federal government to recognize ID.me for its role in preventing large-scale fraud within national systems.

Learn from us: "We’ve been able to scale both our infrastructure and trust. With a platform that’s faster, smarter, and built to handle portable identity at massive scale, we’re one step closer to our goal: a secure, digital way to prove who you are, wherever you need it, that works everywhere you need it." — Kevin Liu, Cloud Platform Architect, ID.me

Manhattan Associates powers more than a billion daily API calls

Who: Manhattan Associates is a global leader in supply chain and omnichannel commerce solutions, offering tools and platforms that reach more than 2 billion people across 20 billion consumer touchpoints.

What they did: Manhattan Associates modernized its Manhattan Active SaaS platform by migrating from legacy Oracle and DB2 systems to Google Cloud databases. Each capability of Manhattan Active now runs as an independent, containerized service orchestrated by Google Kubernetes Engine (GKE). Data flows through Pub/Sub into BigQuery for real-time analytics, while Cloud Logging and Cloud Monitoring deliver observability at scale.

Why it matters: With its new microservices-first design, Manhattan gained the agility to evolve faster and the confidence that mission-critical operations would remain resilient across regions. With Cloud SQL and BigQuery, the company now processes more than a billion daily API calls with average response times of less than 150 milliseconds. This evolution supports hundreds of thousands of monthly active users across tens of thousands of stores and distribution centers. The new platform also created the foundation for Manhattan’s Agentic AI suite, which includes prebuilt agents — like the Intelligent Store Manager and Labor Optimizer — that coordinate real-time decisions across store and distribution center operations.

Learn from us: "Operationally, the platform has become more elastic and efficient. The system automatically handles hundreds of thousands of scaling events per day, ensuring performance remains consistent during peak surges without expensive overprovisioning." — Narayana Reddy Kothapu, Senior Director, Manhattan Associates & Rajkumar Ramani, Technical Director, Manhattan Associates

How a leading consumer insight brand uses Dataproc to hyper-personalise faster

Mon, 06 Apr 2026 16:00:00 +0000

At RVU, we have a clear and vital mission: empower people, transform industries.

For our market-leading home management and switching brands — Confused.com, Uswitch, Tempcover, Money.co.uk, and Mojo Mortgages — transparency and accurate information are everything. Today’s consumer expects more than a simple comparison table; they want personalized recommendations tailored to their unique circumstances.

Delivering on that promise — building a true personalization engine that powers all our brands — requires a data foundation capable of processing massive, complex datasets for sophisticated ML models. Today, our platform powers hundreds of automated personalization campaigns, optimized with billions of data points from across all our brands. We tackled this challenge using the power of Google Cloud and its two solutions for Apache Spark: Dataproc and Google Cloud Serverless for Apache Spark. Together, we’re making our mission a reality.

The high-speed engine for feature engineering

Our relationship with Google Cloud isn’t new. In fact, we've used BigQuery as our unified data platform for over a decade. Coming from a performance marketing background, we’ve always dealt with a lot of data, but we recognized early on that we’re not a digital infrastructure company. Instead, our focus must always be on where the value is. Managed solutions like BigQuery that eliminate infrastructure and capacity headaches were a natural fit from the start.

The key challenge was stitching together a meaningful and coherent picture of customer behavior across our brands — turning countless fragmented interactions into something that genuinely reflects how a user behaves, clicks, and makes decisions. Instead of relying on isolated events and aggregate views, we’ve had to build a platform capable of connecting these signals into a narrative that works for our machine learning models.

Using Dataproc to support this was a gamechanger. The biggest impact has been its role as our core high-speed Spark processing engine, primarily for feature engineering for our ML model development. Feature engineering, which is the crucial process of shaping all that raw customer data for our data science models, is a real value-driver for us. It’s where we have a marked competitive edge.

The result has been a significant leap in our innovation velocity. With Serverless for Apache Spark, we now have the ability to shape our customer data for feature engineering in just a matter of days. Previously, this would have taken weeks. We’ve also dramatically reduced our time-to-market, which also used to take weeks. Now, a new contractor can join the team and deliver a model, including exploratory data analysis and all feature engineering, in only a week and a half. That’s incredibly fast.

Delivering personalized experiences

By improving our speed of innovation, we’re better positioned to deliver a personalized user experience to our customers and partners.

Our hyper-personalization journey accelerated once we moved to Spark. We can now run heavyweight data processing jobs that crunch vast amounts of behavioral and contextual data, allowing us to build models that generate genuinely meaningful predictions.

These models help us understand not just what to say to a customer, but when and how to say it — selecting the right moment and the right channel to deliver personalized insight that genuinely resonates.

Building a future vision

Google’s Data Cloud directly aligns with our culture of prioritizing value, and its impact on our business is profound. I call it the network effect, where everything seamlessly connects within the same ecosystem: Our data resides in BigQuery, our ability to validate, enrich, and transform that data is tied to Dataproc and Serverless for Apache Spark, and our capacity to deploy the ML models spans the network. It’s all co-located and integrated, powering the real-time accuracy of our consumer brands and giving us a competitive advantage.

For our engineers, the big win is the lack of infrastructure they have to deal with. They can press a button that processes all the data in 10 minutes, rather than having to set up a network of clusters and servers and make them talk to each other. It’s incredibly efficient and frees up time for more valuable work like building and iterating data products.

Dataproc has upped our speed, scale, and agility. It also gives us the tools to innovate with AI as we build the future of hyper-personalization. Today, we’re proud to say RVU’s cutting-edge tech and data are helping millions of UK consumers make smarter, more informed decisions, and truly transforming industries.

Inspired by RVU's success? Whether you need persistent clusters with Dataproc or the agility of Serverless Spark, Google Cloud has a managed solution to help you focus on value, not infrastructure. Discover the right Spark on Google Cloud for your use case.

Centralized policy meets distributed logic: Getting to know Eventarc Advanced

Fri, 27 Feb 2026 17:00:00 +0000

Enterprise architects often face a fundamental dilemma: choosing between developer agility and organizational control. Development teams need to move fast and deploy independent microservices without waiting for permission. Security and compliance teams need to be safe, and ensure that data flow is observable and governed by policies.

That’s why we built Eventarc Advanced, a serverless eventing platform and the evolution of Eventarc Standard. Eventarc Advanced provides an improved architectural pattern for the modern cloud, where centralized policy meets distributed logic. By clearly separating the governance layer (the "bus") from the processing layer (the "pipeline"), Eventarc Advanced gives SecOps teams the visibility and control they demand, while freeing developers to choreograph AI agents and build event-driven applications with the autonomy they want. Eventarc Advanced became generally available in August 2025.

In this blog, we take a deeper look at the evolution of integration architectures — from service buses, to microservices, to where we are today — and go into depth with a real-world example. Let’s jump in.

The evolution of integration architectures

To understand the value of this new pattern, it helps to look at where we came from and why previous architecture patterns forced a compromise.

The centralized bottleneck of the Enterprise Service Bus

One early integration architecture approach was the Enterprise Service Bus (ESB), which prioritized centralized control. The ESB emerged to solve the "spaghetti architecture" of point-to-point integrations by providing a centralized communication layer that standardized how disparate systems interact. However, it often introduced serious pitfalls.

The primary issue was what’s referred to as a centralized logic trap. Organizations frequently embedded complex business logic — transformations and orchestration — directly into the governance layer. The resulting middleware layer was opaque, with critical business rules hidden from the developers who owned the services.

Consequently, integration changes typically required the intervention of a central middleware team. Development teams lost autonomy, forced to queue behind integration specialists to ship even minor features, often waiting weeks for updates.

Microservices’ governance gap

To address this, the industry shifted toward microservices (often described as "smart endpoints and dumb pipes"), distributing logic to give teams the autonomy they were looking for. For synchronous traffic (REST, gRPC), tools like API gateways and service meshes restored a layer of governance by enforcing policies like authentication and rate limiting at the infrastructure level.

However, as architectures shifted to Event-Driven Architecture (EDA) for greater resilience and decoupling, a new gap emerged. In a distributed, asynchronous world, centralized control often vanished. This created a governance gap where SecOps teams struggled to maintain order. Three issues emerged to the forefront:

The visibility void: Without a central policy, shadow IT services could silently subscribe to sensitive events without detection.
The policy problem: Enforcing data residency or PII masking is nearly impossible when the broker treats every message as an opaque blob.
The dependency risk: Without clear contracts, changing an event schema risks silently breaking unknown downstream consumers.

A new pattern: Centralized policy, distributed logic

Eventarc Advanced addresses the trade-off between control and speed with a novel architectural pattern: centralized policy meets distributed logic.

Eventarc Advanced maps these distinct responsibilities to two specific architectural resources that each correspond to a distinct role:

The bus: This governance layer is a managed, centralized hub where platform administrators enforce global constraints before events are routed. It synthesizes the centralized routing of the legacy ESB with the modern security architecture of a service mesh. It handles Identity and Access Management (IAM), including content-based access control, to strictly define who can publish, and integrates with VPC Service Controls to prevent data exfiltration.
The pipeline: Think of this distributed, team-owned resource as developers’ integration logic layer. This is where eventing patterns for AI agents and microservices are unlocked, allowing developers to configure event flow and delivery according to their specific business logic. Unlike many service meshes that treat data as opaque bits, the pipeline understands content. Developers can transform events, convert payloads between formats (like JSON to Avro), and configure retry policies and authentication independently.

In other words, by decoupling these duties, Eventarc Advanced provides the control of an ESB with the agility of microservices and the resilience of modern event-driven architectures.

How it works: A retail event mesh example

A typical Eventarc Advanced solution can be implemented with minimal configuration, providing a streamlined experience for both administrative governance and distributed integration logic. To see this model in practice, let's look at a real-world implementation of a retail event mesh.

Imagine an ecosystem at a global retailer with four autonomous teams in charge of the following services:

Commerce
Finance
Logistics
Intelligence (AI Insights Agent)

In a traditional setup, aligning these teams is difficult. The Intelligence team wants access to everything for their models, Finance wants to lock everything down for compliance, Logistics just needs a stable schema to ship boxes, and Commerce needs to roll out new features at a moment’s notice.

The foundation: Built on CloudEvents

Eventarc Advanced uses a data model based on the open CloudEvents standard, which can carry any type of payload. This helps ensure governance and discoverability while retaining flexibility. In our example, before a single event is published, the platform administrator mandates that every message must contain standard attributes and a specific custom extension for governance.

In this example, every event on the bus must carry the following attributes:

type: Standard identifiers for the event instance (e.g., com.retail.order.created)
source: A standard attribute identifying the producer (e.g., //commerce/frontend)
data_sensitivity: A custom extension attribute to categorize risk

In addition, the organization defines three data sensitivity levels:

restricted (High): Severe risk data like Credit Card Tokens or Tax IDs
confidential (Medium): PII like home addresses
general (Low): Safe operational data like Order IDs

This standardized metadata layer allows the bus to enforce policies based on specific attribute names — checking who sent the data (source) and what kind of data it is (data_sensitivity).

The workflow

With this model, the lifecycle of a single order becomes a secure flow where sensitivity changes at every step.

Order placement: The Commerce service publishes order.created to the Bus. The event’s data sensitivity is tagged as general. The AI Insights Agent service subscribes to analyze market trends.
Payment authorization: The Commerce service publishes payment.authorized tagged as restricted (containing a secure token). The Finance service subscribes to capture the token and executes the charge.
Settlement: The Finance service publishes payment.success tagged as general, signaling the transaction is safe to fulfill without exposing financial secrets. Logistics subscribes to ship the box, and Intelligence AI Insights Agent is triggered to evaluate market trends for the next supply chain cycle.
Fulfillment: The Logistics service publishes shipment.ready tagged as confidential (containing the customer phone number). The Logistics own notification pipeline subscribes to it to trigger an SMS notification.

In a legacy architecture, mixing PCI, PII, and operational data on a single bus would be a compliance nightmare. With Eventarc Advanced, it’s a solved problem.

The bus: the governance layer

The platform administrator implements a secure strategy on the bus. Rather than blindly trusting internal services, they enforce global policies that inspect these CloudEvents attributes using fine-grained access control (FGAC).

Enforcing source integrity

To ensure a compromised service cannot spoof events, the bus administrator enforces the producer's identity to match the source attribute.

For example, a bus policy can state that only the principal sa-commerce@retail.com can publish events that match the expression message.source.startsWith("//commerce/"). If the Intelligence AI Insights Agent service tries to publish an event claiming to be from //commerce/payments, the bus rejects the request.

Enforcing a data classification

To ensure every event is categorized, the bus administrator requires that every payload received by the bus includes a valid sensitivity attribute. A bus policy can check that message.data_sensitivity is one of ['general', 'confidential', 'restricted']. This guarantees that the event mesh contains only classified, governance-ready data.

The Pipeline: the logic layer - autonomous team innovation

With the security posture established on the bus, development teams can then use pipelines to solve complex integration challenges entirely within their own domains. Let’s take a look at a few of these challenges.

Schema-aware formats conversion and payload transformation

The Logistics team decides to upgrade their warehouse robots to use high-efficiency protocol buffers. Instead of forcing the Finance team to change their JSON output (which would break other consumers), Logistics configures a transformation step in their own pipeline.

A typical com.retail.payment.success event from Finance arrives as JSON:

code_block: <ListValue: [StructValue([('code', '{\r\n "id": "89d5663e-789e-4d9f-a65f-f7d83742d987",\r\n "source": "//finance/ledger",\r\n "type": "com.retail.payment.success",\r\n "data_sensitivity": "general",\r\n "datacontenttype": "application/json",\r\n "data": {\r\n "order_number": "ORD-2023-8841",\r\n "total_amount": 249.99,\r\n "currency": "USD",\r\n "transaction_id": "tx_77382910",\r\n "status": "SETTLED"\r\n }\r\n}'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe0b8220cd0>)])]>

The warehouse robots service expects a binary Protobuf message:

code_block: <ListValue: [StructValue([('code', 'message PaymentConfirmed {\r\n string order_id = 1;\r\n double insured_value = 2;\r\n string currency_code = 3;\r\n string ledger_reference = 4;\r\n}'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe0b8223850>)])]>

The Logistics team configures their pipeline to accept json as input and output to protobuf. To map the data, they use Common Expression Language (CEL) to configure a transformation:

code_block: <ListValue: [StructValue([('code', '// CEL Transformation to Construct the target Protobuf message\r\n{\r\n "order_id": message.data.order_number,\r\n // 110% of total to cover replacement cost\r\n "insured_value": message.data.total_amount * 1.1,\r\n // Standardize currency to uppercase\r\n "currency_code": message.data.currency.upperAscii(),\r\n "ledger_reference": message.data.transaction_id,\r\n}'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe0b8221c50>)])]>

This transformation not only maps the input but also applies business logic — calculating the insured value and normalizing the currency code. The Logistics team implements this modernization without a single meeting with the Finance team.

Agentic workflows: Filtering and triggering AI agents

Eventarc Advanced enables agentic workflows by allowing pipelines to communicate directly with AI agents using open standard protocols like Agent2Agent (A2A) and Model Context Protocol (MCP), while also offering rich capabilities like filtering to optimize when those agents are invoked.

The Intelligence team uses a pipeline they name ai-insights and the A2A protocol to connect with an AI Insights Agent that proactively analyzes market trends based on placed orders. Because the agent’s processing is resource-intensive, the team uses a filter to only invoke the agent for high-value orders that warrant deeper analysis.

The pipeline filter is configured with the following expression:

code_block: <ListValue: [StructValue([('code', 'message.type == "order.created" && \r\ndouble(message.amount) > 5000.0'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe09a5c13d0>)])]>

When the filter is passed, the pipeline uses a HTTP Message Destination Binding (MDB) expression to directly trigger the agent. By defining a CEL template, the pipeline handles the complexity of constructing a native A2A SendMessage request to the AI strategic insights agent. This allows the agent to receive a conversational prompt derived from technical event data without any manual "glue code":

code_block: <ListValue: [StructValue([('code', '{\r\n "headers": headers.merge({ "Content-Type": "application/json", "A2A-Version": "1.0" }),\r\n "body": {\r\n "jsonrpc": "2.0",\r\n "id": message.id,\r\n "method": "message/send",\r\n "params": {\r\n "message": {\r\n "messageId": message.id,\r\n "role": "ROLE_USER",\r\n "parts": [\r\n { \r\n "text": "Analyze Order " + message.data.order_number + " for market trends." \r\n }\r\n ]\r\n }\r\n }\r\n }\r\n}'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe09a5c0f50>)])]>

A similar prompt message can be crafted for other popular agentic communication protocols like MCP.

This combination of filtering and agentic protocol translation ensures that AI resources are used precisely where they add value. The Intelligence team implements this independently – without writing ingestion code or coordinating with the Commerce or Admin team. The agent can then publish its own strategic recommendation back to the bus, enabling a choreography of AI experts that turns standard cloud events into competitive intelligence.

Advanced API request modeling

When a shipment is ready, the Logistics team uses a pipeline to send an SMS using a legacy gateway API. Integrating with legacy third-party APIs often requires writing "glue code" services just to format requests.

The Logistics team eliminates this maintenance burden by configuring a dedicated pipeline to fully construct the exact request expected by the legacy service.

They use a HTTP Message Destination Binding CEL expression, which standardizes the phone number and maps it to the X-SMS-To HTTP header required by the API. It also construct the SMS text:

code_block: <ListValue: [StructValue([('code', '{\r\n "headers": { "X-SMS-To", \r\n message.data.phone.matches(\'^\\\\+1\') ?\r\n message.data.phone : \r\n \'+1\' + message.data.phone \r\n },\r\n\r\n "body": {\r\n "sms_text": "Order " + message.data.order_id + " shipped!"\r\n }\r\n}'), ('language', ''), ('caption', <wagtail.rich_text.RichText object at 0x7fe09b144810>)])]>

Finally, they configure a robust retry policy (linear backoff, max five attempts) directly on the pipeline, so that temporary network interruptions don't result in lost notifications. In addition to HTTP endpoints, the pipeline supports guaranteed delivery and out-of-the-box authentication for destinations like Cloud Run, Pub/Sub, Bus, Workflows, and over 200 Google services.

The future of agile integration

Eventarc Advanced closes an important gap in event-driven architectures: It brings the same level of maturity to asynchronous communication by introducing the pattern of centralized policy, distributed logic.

For the Platform team, Eventarc Advanced provides assurance that a bus can strictly enforce integrity and confidentiality on every message, bringing "service-mesh-like" security to the event layer.
For the developer, it restores autonomy. The pipeline allows teams to filter, transform, convert, and route events to fit their specific needs, enabling them to treat events as first-class products rather than opaque artifacts.

This architecture lays the foundation for the next generation of intelligent applications. A secure, typed, and trustworthy event mesh can serve as the backbone for generative AI agents and real-time analytics, allowing you to safely expose business context to the systems that need it most.

Get started

Don't let governance slow down your innovation. Here are some Eventarc Advanced resources to get you on your way:

Learn more: Dive into the full capabilities of the Bus and Pipeline in the Eventarc Advanced documentation.
Get hands-on: Deploy the "Retail Event Mesh" scenario yourself and explore enterprise patterns with our Quickstarts and Tutorials.
Start building: Go to the Google Cloud console to configure your first bus and pipeline today.
Let's talk: Have a complex enterprise use case? Contact Google Cloud Sales to discuss how Eventarc Advanced fits into your broader integration strategy.

Cool stuff Google Cloud customers built, Feb. edition: Telco data reinvention; Golden State’s “G.O.A.T.T.”; John Lewis explores DORA

Thu, 26 Feb 2026 17:00:00 +0000

AI and cloud technology are reshaping every corner of every industry around the world. Without our customers, there would be no Google Cloud, as they are the ones building the future on our platform. In this regular round-up, we dive into some of the exciting projects redefining businesses, shaping industries, and creating new categories.

For our latest edition, we explore a new data approach for Vodafone and Fastweb; evaluating John Lewis Partnership’s developer platforms; the Golden State Warrior’s AI playbook; healthy, stable networks at Hackensack Meridian Health; and Ab Initio brings better context to data for AI.

Be sure to check back next year to see how more industry leaders and exciting startups are putting Google Cloud technologies to use. And if you haven’t already, please peruse our list of 1,001 real-world gen AI use cases from our customers.

Fastweb + Vodafone reimagined data workflows

Who: Following the acquisition of Vodafone Italy by Swisscom in 2025, these leading European telecom providers wanted to rethink how they serve customers and deliver timely, personalized experiences across mobile, broadband, and digital channels.

What they did: Both companies had already begun modernizing customer data workflows with BigQuery, but combining ecosystems exposed certain limits of the existing setup. In order to give every channel real-time access to accurate customer data, they implemented Spanner as a service and governance layer, delivering low-latency reads, horizontal scalability, high availability, and a fully managed environment with zero ops overhead. The team is also using Gemini to generate clear documentation directly from the code, which saves hours of manual work.

Why it matters: Using Spanner Graph allowed the organization to map lineage in a way that reflects how its platform actually works: which tables drive specific jobs, how transformations cascade, and where dependencies sit. Call centers now see more complete, up-to-date customer information, digital channels can rely on consistent data without custom integrations, and partners can access what they need with low latency through Apigee.

Learn from us: “Rebuilding our Customer 360 platform with Google Cloud services has already changed how Fastweb + Vodafone works. Workflow monitoring is simpler, pipelines are leaner, and real-time serving is now the norm. ” – Vincenzo Forciniti, IT AI Adoption & Platform Engineering Lead, Fastweb + Vodafone

John Lewis measures the value of its developer platform

Who: The John Lewis Partnership is a major UK retailer operating John Lewis department stores and Waitrose supermarkets. To power their digital transformation, they built the John Lewis Digital Platform (JLDP) to support dozens of product teams building high-quality software for johnlewis.com.

What they did: Moving beyond simple usage metrics, John Lewis developed a sophisticated, multi-stage approach to measuring the real value of their platform. They transitioned from initial speed-based metrics (like "Onboarding Lead Time") to a comprehensive model using DORA metrics and subjective engineer feedback via the DX platform. This included a custom "Technical Health" feature that uses small, automated jobs to monitor more than 35 health measures — such as Kubernetes best practices, security, and operational readiness — providing teams with real-time "traffic light" indicators of their service health.

Why it matters: By focusing on value rather than just activity, John Lewis ensured the platform was actually reducing friction for developers rather than just being a mandatory tool. Their automated Technical Health checks allow product teams to manage technical debt and security vulnerabilities proactively. This approach has decoupled centralized operations teams from individual services, leading to faster incident resolution (MTTR), fewer outages, and significant cost savings.

Learn from us: "Measurement is a journey, not a destination. Start by measuring something meaningful to your stakeholders, but be prepared to adapt as your platform evolves. The things that mattered when you were proving out the platform's viability are unlikely to be what are important several years later when your features are mature." – Alex Moss, Principal Platform Engineer, John Lewis Partnership

Hackensack Meridian Health de-risks network migration using VPC Flow Logs

Who: Hackensack Meridian Health is a leading not-for-profit healthcare organization and the largest hospital system in New Jersey. System reliability is a cornerstone value for HMH as they manage a vast network of hospitals, urgent care centers, and physician practices.

What they did: Preparing for a large-scale migration to a new Google Cloud network design, Hackensack Meridian Health used VPC Flow Logs and Flow Analyzer to eliminate the "black box" of hybrid traffic. By enabling logs on their Cloud Interconnect VLAN attachments, they captured granular telemetry — including source/destination IPs, ports, and protocols. They then exported this data to create a visual "who-is-talking-to-what" map. This allowed them to identify critical traffic patterns between on-premises data centers and specific Google Cloud regions, VPCs, and applications.

Why it matters: In a healthcare environment, even minor network disruptions can have major consequences. By mapping traffic proactively, Hudson Meridian Health pinpointed exactly which moments in the cutover carried the highest risk. This preparation allowed them to detect a migration issue in just three minutes and resolve it within five — a process that previously could have taken hours. Beyond migration, this level of visibility enables the organization to better manage capacity planning, cost attribution, and security compliance across their hybrid infrastructure.

Learn from us: "Getting a clear picture of our interconnect traffic always felt like a black box. Enabling VPC Flow Logs and feeding it into Flow Analyzer finally gave us the map we needed. Identifying those critical traffic flows before we changed any routes was key to de-risking the entire migration." — Randall Brokaw, Cloud Engineering Manager, Hackensack Meridian Health

The Golden State Warriors’ AI-powered back office

Who: The Golden State Warriors are one of the NBA’s most successful modern franchises. Behind their on-court wins are a specialized operations team who run what might be called organization’s "G.O.A.T.T." (Greatest of All-Time Technologies), a data and AI platform that helps drive game-time insights, trading decisions, and fan experience enhancements.

What they did: The Warriors transitioned from a "gut-feeling" culture to an "analytics-first" strategy by building an internal "digital brain" on Google Cloud. Using BigQuery and Gemini, the team now automates complex workflows that previously took hours, such as generating pre-game scouting reports. They use machine learning to run thousands of trade simulations that prioritize "team fit" over raw individual stats and employ computer vision to track the "shot quality" of every attempt in the NBA. On the business side, they built a content recommendation engine using the Discovery API to deliver personalized digital experiences to their global fan base.

Why it matters: This AI-driven approach narrows the decision tree for leadership, allowing them to focus human expertise on the most viable options. By automating the “science” of data processing, coaches and scouts have more time for the "art" of face-to-face training, planning, and player development. This integration has not only influenced on-court strategy — like the three-point revolution — but has also improved business efficiency, with employees now proactively bringing AI-driven ideas to the IT team rather than waiting for top-down mandates.

Learn from us: "You can never reach a point where either humans or machines are making all the decisions. The sweet spot is finding that middle ground where intuition and data converge on the same conclusion. Data helps us narrow our decision tree before we even start evaluating specific options." — Nick Manning, Senior Director of Consumer Products & Emerging Technology, Golden State Warriors

Ab Initio unlocks enterprise data for the agentic AI era

Who: Ab Initio is an enterprise software company specializing in high-volume data integration and governance. Their platform is trusted by large-scale organizations to manage complex data lifecycles across hybrid and multi-cloud environments.

What they did: To solve the challenge of grounding AI agents in accurate data, Ab Initio partnered with Google Cloud to integrate its data fabric with BigQuery, Dataplex Universal Catalog, and Gemini. They launched a suite of more than 500 metadata and data connectors that bridge the gap between legacy systems (like mainframes, COBOL, and SAS) and modern cloud environments. This integration provides field-level, end-to-end lineage, allowing Gemini to access well-documented, "AI-ready" data regardless of where it resides.

Why it matters: AI agents are only as effective as the data they can access. By using Ab Initio as a "neutral hub," enterprises can federate data from on-premises and multi-cloud sources into a single unified layer without moving the data itself. This provides the rich semantic context and lineage needed for Gemini to perform grounded, explainable reasoning. For businesses, this means faster transition from experimental AI to production-ready agentic workflows that are auditable, compliant, and capable of making complex, automated decisions.

Learn from us: "Agentic AI requires trusted, AI-ready data and metadata. Understanding the origin, quality, and meaning of information matters as much as the data itself. Gemini serves as a key component of the agentic layer, using this context to make decisions that are explainable and auditable." — Scott Studer, Head of Development, Ab Initio & Chai Pydimukkala, Data Governance, Sharing & Integration Product Lead, Google Cloud

Cool stuff Google Cloud customers built, Dec. edition: AI for better toys, reliable mapping tech, Gemini stumps an all-star & more

Wed, 31 Dec 2025 16:00:00 +0000

For our latest edition, we look into how Waze made its network more reliable; NBA superstar Stephen Curry gets quizzed by Gemini; a financial market transformation at CME Group; a multi-agent business forecasting platform from AppOrchid; Mattel crunches customer feedback with AI; VMO2 uses decentralized contracts for reliable data; Mercado Libre’s strategic use of Spanner; and how Ericsson enhances data governance.

Waze keeps traffic flowing with Memorystore

Who: Waze (a division of Google parent company Alphabet) is a community-driven, crowd-sourced navigation app with tens of millions of users who share real-time data to provide optimal driving routes, traffic updates, and alerts for hazards, police, and more.

What they did: Waze depends on vast volumes of dynamic, real-time user session data to power its core navigation features, but scaling that data to support concurrent users worldwide required a new approach. Their team built a centralized Session Server backed by Memorystore for Redis Cluster, a fully managed service with 99.99% availability that supports partial updates and easily scales to Waze’s use case of over 1 million MGET commands per second with ~1ms latency.

Why it matters: Moving from Memcached’s 99.9% SLA to Memorystore for Redis Cluster’s 99.99% means higher availability and resiliency from the service. And because Memorystore for Redis supports partial updates, Waze can change individual fields within a session object rather than rewriting the entire record. That reduces network traffic, speeds up write performance, and makes the system more efficient overall.

Learn from us: “Real-time data drives the Waze app experience. Our turn-by-turn guidance, accident rerouting, and driver alerts depend on up-to-the-millisecond accuracy. But keeping that experience seamless for millions of concurrent sessions requires robust and battle hardened infrastructure that is built to manage a massive stream of user session data.” – Eden Levin, Waze BE infrastructure developer & Yuval Kamran Waze site reliability engineer

What Stephen Curry learned from a custom Gemini agent

Who: Stephen Curry is arguably of the greatest three-point shooter of all-time in the NBA — as well as Google’s performance advisor and an all-around stats-obsessive.

What they did: For a special engagement with Curry, the Google Cloud team wanted to showcase the power of Gemini for creative thinking, analysis, and data mining. They took every regular season, play-in, and playoff game from Curry’s career (through the end of the 2024-2025 season) and input the data into a custom-built agent using Google Cloud’s Agent Development Kit and Gemini APIs.The system could then be queried for obscure stats, to see if the team could stump Curry and teach him more about his game.

Why it matters: For example, it found that his three-point shooting percentage after more than seven dribbles, with a minimum 105 attempts was 40.2%, and how many points Curry generated for his teammates off of screens since 2013: 1,105. Instead of countless hours of manual research, the team got query results in less than a minute. Some queries were so obscure, the team wouldn’t have reached a valid answer without the ability of the agent to analyze the rich data.

Learn from us: “Gemini is going to be in my head this year, cause I'm going to be looking at all these details.” – Stephen Curry, Golden State Warriors point guard and 4x NBA champ

How CME Group builds a faster, smarter exchange

Who: CME Group has evolved from a nineteenth-century commodities exchange into one of the most advanced financial market infrastructures in the world. To support real-time trading and risk management at a global scale, the company launched a strategic partnership with Google Cloud.

What they did: By migrating to Cloud SQL and adopting AI-powered insights, CME Group empowered developers, paid down technical debt, and unlocked new opportunities for data-driven innovation across financial markets.

Why it matters: Cloud SQL has given CME a foundation for increased developer and team agility. Fewer performance issues mean more time focused on innovation: expanding CME’s analytics capabilities, accelerating AI initiatives, and exploring new ways to commercialize data responsibly. When teams stopped chasing outages, they unlocked more time to take bigger bets and build the future.

Learn from us: “With Cloud SQL, we’ve found a way to keep our data layer as fast and dependable as the markets we serve. Cloud SQL gives our teams real-time visibility into what’s happening inside the database. When an application slows, we can identify the root cause in minutes instead of hours. Those insights are built into the platform, which means we don’t need custom tooling or manual analysis to keep operations steady.” – Kristofer Shane Sikora, Executive Director, Cloud Data Engineering, CME Group

AppOrchid’s multi-agent system for superior business forecasting

Who: App Orchid is an enterprise AI builder and a leader in making data actionable with AI, with a mission to make AI a force for good. Their goal is to empower every employee with trusted, understandable, and accessible data.

What they did: The business forecasting agent is actually built on the foundation of two powerful, specialized AI agents: a prediction agent built by Google Cloud and App Orchid’s Data Agent offering. These agents work in concert to solve complex business problems, acting as complementary specialists. App Orchid’s agent possesses unparalleled understanding of an enterprise's past and present, while Google’s agent brings world-class capabilities in predicting the future.

Why it matters: Adopting a multi-agent approach provides clear, tangible advantages that directly address the forecasting problems that often plague businesses, including improved accuracy; increased operational efficiency; faster insights; and reduced costs and increased revenue; and greater agility and adaptability. Neither of the underlying agents could achieve these results on their own, but working together, this agent is more than the sum of its subagents.

Learn from us: “As the agentic era gets underway, it is evolving quickly. Our multi-agent approach demonstrates both how true agentic systems are most successful when multiple agents are at play, and the importance of finding strong partners with distinct capabilities to help build and assemble these agentic systems.” – Brian Mills, Director, Enterprise AI, Google Cloud & Taka Shinagawa, Gen AI Field Solution Architect, Google Cloud

How Mattel uses AI for real-time product updates

Who: Since its humble beginnings in a garage in 1945, Mattel has consistently been reshaping play for children and families across the globe with iconic franchises like Barbie, Hot Wheels, and Fisher-Price.

What they did: To improve its understanding of consumer sentiment, Mattel developed an AI-powered feedback classification system, which can analyze millions of customer interactions from a diverse range of sources in a matter of seconds. At its core, the system relies on BigQuery for storing and efficiently processing its massive customer datasets and then utilizes Vertex AI and Google’s multimodal Gemini models to refine and train the sophisticated consumer feedback model.

Why it matters: Already, the new AI-powered system has delivered significant wins, delivering a staggering 100x boost in data processing capacity and reducing analysis times from a month to a single minute. By automating the analysis of many processes, analysts are now freed from the noise of everyday tasks, enabling them to focus on deeper research across the company’s iconic portfolio brand.

Learn from us: “Our big motto is ‘From months to minutes,’ but it’s real. We were literally spending months-worth of analysis and just getting data into the place that an analyst could tally up all the sentiment — and now it’s just at our fingertips.” – Shaun Applegate, Director of Product Quality Analytics, Mattel

Virgin Media O2 uses data contracts for scalable AI products

Who: Virgin Media O2 is one of Europe’s largest telecommunications and media providers, with 45.8 million broadband, mobile, phone, and home subscribers across the UK. To build AI products that are adaptable and data-driven, they needed a decentralized system that internal customers could count on for clean, reliable data.

What they did: New decentralized data contracts, built with Dataplex, serve as the data quality and assurance layer for VMO2’s data products; these ensure every dataset they publish is reliable, documented, and ready for consumption. Defined at the asset level, such as individual BigQuery tables or Google Cloud Storage buckets, data contracts are redefining how VMO2 manage and share data, enabling the creation of trusted and scalable AI products across their data mesh.

Why it matters: The power of this approach lies in moving beyond static documentation. Because they are machine-readable, data contracts become living guarantees with continuous enforcement and real-time validation directly within data pipelines. This proactive monitoring allows teams to detect schema changes or SLA breaches early, transforming data quality from a reactive fix into a scalable, automated mechanism.

Learn from us: “By operationalizing trust through data contracts, we are fostering a culture of shared responsibility and data-first thinking. This federated model does more than simply fix pipelines; it builds the trusted foundation needed to scale next-generation AI. It ensures that the resilient AI tools empowering our teams are built on data that is reliable, consistent, and well-defined.” – Chandu Bhuman, Head of Data Strategy, Cloud & Engineering, Virgin Media O2 & Dženan Softić, Data & AI Architect, Google Cloud

Inside Mercado Libre's multi-faceted Spanner architecture

Who: Mercado Libre, an e-commerce and fintech pioneer across Latin America, operates at a staggering scale, demanding an infrastructure that's not just resilient and scalable, but also a catalyst for rapid innovation.

What they did: At the heart of Mercado Libre's strategy is Fury, an in-house middleware platform designed to abstract away the complexities of various backend technologies, providing developers with standardized, simplified interfaces to build applications. Spanner provides Fury with an always-on, globally consistent, multi-model database with virtually unlimited scale. By designating Spanner as a choice within Fury, Mercado Libre ensures that applications built on the platform using Spanner stay consistent globally, scale without breaking, and rarely go down.

Why it matters: The strategic adoption of Spanner, amplified by internal platforms like Fury and sophisticated data workflows, has yielded significant benefits, including: significant cost savings and low total cost of ownership; business impact and agility for developers; and low operational overhead thanks to automation.

Learn from us: “Mercado Libre's adoption of Spanner demonstrates how to use a powerful, globally consistent database not just for its core capabilities, but as a strategic enabler for developer productivity, operational efficiency, advanced analytics, and future AI ambitions.” – Pablo Leopoldo Arrojo, Software Technical Leader, Mercado Libre

Ericsson achieves data integrity and superior governance with Dataplex

Who: Ericsson is one of the world's leading providers of telecommunications and networking technology and solutions. Its Managed Services unit provides network operations and optimization, including field operations, for various telecom and enterprise customers, including outsourcing network performance management, future provisioning, network vulnerability management, and network energy infrastructure management.

What they did: To power the future of its autonomous network operations and deliver on its strategic priorities across a global network of more than 710,000 sites, Ericsson's Managed Services has been on a transformative data journey with governance at the center of its strategy. Ericsson moved from foundational practices to a sophisticated, business-enabling data governance framework using the Dataplex Universal Catalog — turning data from a simple resource into a strategic asset.

Why it matters: With Dataplex as the governance foundation, Ericsson began implementing the core pillars of its governance program, moving from manual processes to an automated, intelligent data fabric. More specifically, Ericsson established a unified business vocabulary within Dataplex, which helped eliminate ambiguity and ensure their teams — from data scientists to data analysts — were speaking the same language.

Learn from us: “Governance is a value enabler, not a blocker. A modern data governance program should focus on business enablement first, driving value and innovation in order to complement policies, rules and risk management. Also remember this work is a journey, not a destination. Be prepared to fail fast, learn, and adapt. The landscape is constantly changing at breakneck speed.” – William McCann Murphy, Head of Data Authority, Ericsson & Akanksha Bhagwanani, EMEA Data Analytics Solution Lead, Google Cloud

How we built a multi-agent system for superior business forecasting

Thu, 11 Dec 2025 17:00:00 +0000

In today's dynamic business environment, accurate forecasting is the bedrock of efficient operations. Yet, businesses across all industries grapple with the constant challenge of predicting future demand, resource needs, and market trends. Nor is this an abstract problem; the cost of miscalculation can be substantial, leading to wasted resources and missed opportunities.

Imagine a large retail chain struggling to predict seasonal demand for its popular clothing lines. A miscalculation means either mountains of unsold inventory and costly markdowns or constant stock-outs that lead to lost sales and frustrated customers. Or consider a manufacturer trying to optimize the procurement of raw materials. Inaccurate forecasts force them into a reactive cycle of expensive rush orders and production delays, or they see their capital tied up in slow-moving inventory.

Whether it’s outdated processes, siloed systems, or missing data semantics slowing down your forecasts — and your business — there’s now a better way. A new era of AI-powered enterprise intelligence means we can move beyond reactive measures and achieve new levels of foresight. Google Cloud and App Orchid have now developed a novel multi-agent application for business forecasting that helps transform the predictive challenges of the past into strategic advantages.

In this post, we will outline the elements of our multi-agent system, the benefits of this innovative approach to agentic frameworks, and how others can benefit from not only our offering but rethink how they create their own.

Transforming enterprise intelligence with AI agents

App Orchid is a leader in making data actionable with AI, with a mission to make AI a force for good. The goal is to empower every employee with trusted, understandable, and accessible data. While enterprise data is now a mission-critical asset for organizations, it’s often underutilized, difficult to access, and buried under layers of complexity. To help, App Orchid partnered with Google to build a multi-agent application that provides a new level of precision in operational forecasting.

This innovative solution combines two powerful, specialized AI agents: a prediction agent built by Google Cloud and App Orchid’s Data Agent offering. These agents work in concert to solve complex business problems, acting as complementary specialists — each an expert in its own domain. App Orchid’s agent possesses unparalleled understanding of an enterprise's past and present, while Google’s agent brings world-class capabilities in predicting the future.

Adopting a multi-agent approach provides clear, tangible advantages that directly address the forecasting problems that often plague businesses, including:

Improved accuracy: Achieve a level of forecasting precision that was previously unattainable, reducing costly errors.
Increased operational efficiency: Automate and streamline the complex processes of data preparation and prediction, freeing up valuable human resources.
Faster insights: Gain real-time, actionable insights, enabling quicker and more informed decision-making.
Reduced costs and increased revenue: Directly impact the bottom line by minimizing inventory waste, reducing stock-outs to maximize sales, and optimizing resource allocation.
Greater agility and adaptability: Rapidly adapt to market shifts and unforeseen disruptions with agile forecasting capabilities.

How it works: Three powerful agents, one seamless solution

To better understand the power of this new agentic framework, it's first essential to understand how these two AI agents work together as complementary specialists under the direction of a third, orchestrator agent.

1. Google prediction agent - The forecasting powerhouse

The prediction agent, which is primarily the custom engineering work of Google Cloud, is the system’s window to the future. It takes rich, contextualized historical data and applies Google's state-of-the-art predictive models to generate highly accurate and actionable forecasts. The agent utilizes specialized foundation models like TimesFM, which is pre-trained on billions of data points for time-series forecasting, and the Population Dynamics Foundation Model (PDFM), which analyzes geospatial data to understand demographic similarities. By combining these powerful models, the Google prediction agent helps businesses anticipate future demand and identify market trends with a new level of precision.

2. App Orchid Data Agent - The enterprise intelligence data expert

Accurate predictions depend on high-quality, AI-ready data, which is where App Orchid’s Data Agent excels. This agent acts as the intelligent query engine for your enterprise data, connecting even disparate and siloed information to draw insights and offer informed feedback. The Data Agent utilizes the Google Cloud Cortex Framework for a unified view of business data and then applies App Orchid’s own powerful semantic-layer technology, which closes the "AI context gap" by transforming raw, siloed data into a unified, trustworthy knowledge foundation that grounds AI models for accuracy and scalability.

Together, this creates a "smart data layer" within the Data Agent that maps an organization’s entire data landscape into a unified knowledge graph. This allows the agent to understand the unique context, relationships, and terminology specific to a business — from internal acronyms to complex operational data — and deliver the comprehensive, time-series datasets that are essential for producing accurate predictions.

3. The combined business forecasting agent

At the heart of the solution is a unified business forecasting agent, which brings together the capabilities of our unique prediction and data agents in a discrete instance for the user. While our multi-agent system delivers an automated, end-to-end tool for next-generation forecasting, it’s the business forecasting agent that the user interacts with — few actually realize there are multiple agents working behind the scenes, nor need to know.

The process begins by having the forecasting agent function as an orchestrator of user queries, directing the App Orchid Data Agent to construct a unified, trustworthy knowledge foundation; this is a crucial step for eliminating data silos and prepares the high-context, time-series data required for accurate AI. This smart data layer is then passed to the Google prediction agent, which applies its specialized foundation models to project future outcomes with impressive levels of precision.

The final result is a single, comprehensive forecast delivered back to the user via the orchestrator agent. And by automating the labor-intensive processes of data wrangling and prediction execution, the combined solution empowers business leaders to make instant, highly accurate decisions that directly reduce costly markdowns, prevent stock-outs, maximize sales, and efficiently optimize resource allocation.

The technical glue: A2A Protocol, Google’s Agent Development Kit (ADK), and Gemini

The seamless collaboration between these two distinct AI agents is made possible by three key technologies: the Agent-to-Agent (A2A) Protocol, Google's Agent Development Kit (ADK), and Google’s industry-leading family of multimodal Gemini models.

The A2A Protocol enables AI agents, even those built by different teams and organizations, to discover, securely communicate, and collaborate across various systems. This allows developers to unite agents from multiple platforms, improving modularity, reducing vendor lock-in, and speeding up innovation. While Model Context Protocol (MCP) allows developers to connect data and APIs to agents, the agents themselves need a communication layer. The A2A protocol enables the bi-directional agentic communication needed to achieve multi-agent systems.

The Google ADK provides a robust framework for building sophisticated, scalable multi-agent applications. It supports both code-first development (Python and Java SDKs) and a no-code (YAML-based) development to define agent behavior and orchestrate workflows. The ADK also utilizes MCP Tools, which provide a standardized interface for agents to interact with external systems.

For instance, the MCP Toolbox for Databases provides out-of-the-box support for agents to easily access and query data from a variety of sources, including BigQuery. Agents built with the ADK can be deployed across virtually any environment — from a fully managed, enterprise-grade runtime like Vertex AI Agent Engine to Google Cloud services like Compute Engine, Cloud Run, or Google Kubernetes Engine (GKE), other cloud providers, or on-premises infrastructure.

At the heart of both agents' intelligence are Google's Gemini models, which are engineered for sophisticated reasoning and lead the industry on long context performance, with many models offering context windows of 1 million tokens or more. This massive context window is critical: it enable the data agent to understand natural language queries, enterprise data, and its underlying schema and relationships simultaneously, while enabling the prediction agent to analyze complex historical data and grasp the nuances of forecasting tasks.

This holistic understanding is what makes it possible for both agents to break down complex problems, identify subtle trends, and follow multi-step instructions without losing the initial context. Additionally, native tool use allows the agents to proficiently interact with external systems — whether querying a database or invoking a predictive model — and interpret the outputs to fulfill user requests.

Solution architecture: A secure, managed platform for AI agents

The Google and App Orchid multi-agent application is accessible through Gemini Enterprise, which provides a fully managed platform for organizations to discover, manage, and interact with AI agents, featuring enterprise-grade security, data privacy, and governance. The user-facing agent is deployed on Vertex AI Agent Engine and registered within this secure environment, providing authorized users with a simple entry point to the powerful multi-agent capabilities while ensuring that all interactions are grounded in the company's private data and adhere to its security and compliance policies.

From an architectural perspective, the business forecasting agent acts as the orchestrator agent of this multi-agent system, managing and directing the entire workflow between the two specialized agents. It uses the A2A protocol to pass instructions and data back and forth, hiding the underlying complexity from the user.

For example, when a user asks the business forecasting agent to predict revenue per marketing channel, the agent:

Calls upon the App Orchid Data Agent to retrieve historical sales data.
Passes that information to the Google prediction agent to run its models and generate the forecast.
Receives the prediction data back and then summarizes it for the user.

The Future is Collaborative

As the agentic era gets underway, it is evolving quickly. Our multi-agent approach demonstrates both how true agentic systems are most successful when multiple agents are at play, and the importance of finding strong partners with distinct capabilities to help build and assemble these agentic systems.

The partnership between App Orchid and Google is a testament to the power of collaboration in the AI space. By combining App Orchid's deep understanding of enterprise data and Google's global leadership in AI, this solution is greater than the sum of its parts. With the complementary strengths of each agent, enabled by the ADK and A2A Protocol, businesses can achieve levels of forecasting accuracy that were previously unattainable at the speed and ease agents can offer.

This innovative multi-agent system not only optimizes business processes but also paves the way for a future where AI agents from across the industry can collaborate to solve the world's most complex challenges.

_{The Google Cloud team would like to thank App Orchid’s CTO Ravi Bommakanti for his contributions to this project.}

Is your DR plan just wishful thinking? Prove your resilience with chaos engineering

Mon, 08 Dec 2025 17:00:00 +0000

When was the last time you knew — not just hoped — that your disaster recovery plan would work perfectly?

For most of us, the answer is unclear. Sure, you may have a DR plan, a meticulously crafted document stored in a wiki or a shared drive, that gets dusted off for compliance audits or the occasional tabletop drill. You assume its procedures are correct, its contact lists are current, and its dependencies are fully mapped, and you certainly hope it works.

But hope is not a strategy.

Why wouldn’t it work? One problem is that systems are rarely static anymore. In a world where you deploy new microservices dozens of times per day, make constant configuration changes, and maintain an ever-growing web of third-party API dependencies, the DR plan you wrote last quarter is probably just as useful as one from 10 years ago.

And if the failover does work, will it work well enough to meet the promises you've made to your customers (or board of directors or regulators)? When a key component fails, could you still even meet your target availability and latency targets, a.k.a., your Service Level Objectives (SLOs)?

So, how do you close this gap between your current aspirational DR plan and a DR plan that you actually have confidence in? The answer isn't to write more documents or run more theatrical drills. The answer is to stop assuming and start proving.

This is where chaos engineering comes in. Unlike what the name might imply, chaos engineering isn’t a tool for recklessly breaking things. Instead, it’s a framework that provides data-driven confidence in your SLOs under stress. By running controlled experiments that simulate real-world disasters like a database failover or a regional outage, you can quantitatively measure the impact of those failures on your systems’ performance. Chaos engineering is how you transform your DR hypotheses into a proven method to ensure resilience. By validating your plan through experimentation, you create tangible evidence, verifying that your plan will safeguard your infrastructure and keep your promises to customers.

Demystifying chaos engineering

In a nutshell, chaos engineering is the practice of running controlled, scientific experiments to find weaknesses in your system before they cause a real outage.

At its core, it’s about building confidence in your system’s resilience. The process starts with understanding your system's steady state, which is its normal, measurable, and healthy output. You can't know the true impact of a failure without first defining what "good" looks like. This understanding allows you to form a clear, testable hypothesis: a statement of belief that your system's steady state will persist even when a specific, turbulent condition is introduced.

To test this hypothesis, you then execute a controlled action, which is a precise and targeted failure injected into the system. This isn't random mischief; it's a specific simulation of real-world failures, such as consuming all CPU on a host (resource exhaustion), adding network latency (network failure), or terminating a virtual machine (state failure). While this action is running, automated probes act as your scientific instruments, continuously monitoring the system's state to measure the effect.

Together, these components form a complete scientific loop: you use a hypothesis to predict resilience, run an experiment by applying an action to simulate adversity, and use probes to measure the impact, turning uncertainty into hard data.

Using chaos to validate disaster recovery plans

Now that you understand the building blocks of a chaos experiment, you can build the bridge to your ultimate goal: transforming your DR plan from a document of hope into an evidence-based procedure. The key is to stop seeing your DR plan as a set of instructions and start seeing it for what it truly is: a collection of unproven hypotheses.

When you think about it, every significant statement in your DR document is a claim waiting to be tested. When your plan states, "The database will failover to the replica in under 5 minutes," that isn't a fact, it's a hypothesis. When it says, "In the event of a regional outage, traffic will be successfully rerouted to the secondary region," that's another hypothesis. Your DR plan is filled with these critical assumptions about how your system should behave under duress. Until you test them, they remain nothing more than educated guesses.

Chaos experiments are the ultimate validation tools, live-fire drills that put your DR hypotheses to a real, empirical test. Instead of just talking through a scenario, you use controlled actions to safely and precisely simulate the disaster. You're no longer asking "what if?"; you're actively measuring "what happens when."

For example, imagine you have a DR plan for a regional outage. When you adopt chaos engineering, you break down that plan into a hypothesis and an experiment. For example:

The hypothesis: "In case our primary region us-central1 becomes unreachable, the load balancers will failover all traffic to us-east1 within 3 minutes, with an error rate below 1%."
The chaos experiment: Run an action that simulates a regional outage by injecting a "blackhole" that drops all network traffic to and from us-central1 for a limited time. Your probes then measure the actual failover time and error rates to validate the hypothesis.

In other words, by applying the chaos engineering methodology, you systematically move through your DR plan, turning each assumption into a proven fact. You're not just testing your plan; you're forging it in a controlled fire.

Connecting chaos readiness to your SLOs

Beyond simply proving system availability, chaos engineering builds trust in your reliability metrics, ensuring that you meet your SLOs even when services become unavailable. An SLO is a specific, acceptable target level of your service's performance measured over a specified period that reflects the user's experience. SLOs aren't just internal goals; they are the bedrock of customer trust and the foundation of your contractual service level agreements (SLAs).

A traditional DR drill might get a "pass" because the backup system came online. But what if it took 20 minutes to fail over, during which every user saw errors? What if the backup region was under-provisioned, and performance became so slow that the service was unusable? From a technical perspective, you "recovered." But from a customer's perspective, you were down.

A chaos experiment, however, can help you answer a critical question: "During a failover, did we still meet our SLOs?” Because your probes are constantly measuring performance against your SLOs, you get the full picture. You don't just see that the database failed over; you see that it took 7 minutes, during which your latency SLO was breached and your error budget was completely burned. This is the crucial, game-changing insight. It shifts the entire goal from simple disaster recovery to SLO preservation, which is what actually determines if a failure was a minor hiccup or a major business-impacting incident. It also provides the data necessary to set goals for system improvement. So the next time you run this experiment, you can measure if and how much your system resilience has improved, and ultimately if you can maintain your SLO during the disaster event.

Build a culture of confidence

The journey to resilience doesn't start by simulating a full regional failover. It starts with a single, small experiment. The goal is not to boil the ocean; it's to build momentum. Test one timeout, one retry mechanism, or one graceful error message.

The biggest win from your first successful experiment won't be the technical data you gather. It will be the confidence you build. When your team sees that they can safely inject failure, learn from it, and improve the system, their entire relationship with failure changes. Fear is replaced by curiosity. That confidence is the catalyst for building a true, enduring culture of resilience. To learn more and get started with chaos engineering, check out this blog and this podcast. And if you’re ready to get started, but unsure how, reach out to Google Cloud professional services to discuss how we can help.

Streamline the design and deployment of application infrastructure with Application Design Center, now GA

Mon, 08 Dec 2025 17:00:00 +0000

Earlier this year, we unveiled a big investment in platform and developer team productivity, with the launch of Application Design Center, helping them streamline the design and deployment of cloud application infrastructure, while ensuring applications are secure, reliable, and aligned with best practices. And today, Application Design Center is generally available.

We built Application Design Center to put applications at the center of your cloud experience, with a visual, canvas-style and AI-powered approach to design and modify Terraform-backed application templates. It also offers full lifecycle management that’s aligned with DevOps best practices across application design and deployment.

Application Design Center is a core component of our application-centric cloud experience. When you use Application Design Center to design and deploy your application infrastructure, your applications are easily discoverable, observable, and manageable. Application Design Center works in concert with App Hub to automatically register application deployments, enabling a unified view and control plane for your application portfolio, and Cloud Hub, to provide operational insights for your applications.

“Google Application Design Center is a valuable enabler for Platform Engineering, providing a structured approach to harmonizing resource creation in Google Cloud Platform. By aligning tools, processes, and technologies, it streamlines workflows, reducing friction between development, operations, and other teams. This harmonization enhances collaboration, accelerates delivery, and ensures consistency across Google Cloud environments.” - Ervis Duraj, Principal Engineer, MediaMarktSaturn Technology

The gateway to an app-centric cloud

Our goal with Application Design Center is for you to innovate more, and administer less. It consists of four key elements to help you minimize administrative overhead and maximize efficiency, so you can design and deploy applications with integrated best practices and essential guardrails. Let’s take a closer look.

1. Terraform components and application templates
Develop applications faster with our growing library of opinionated application templates. These provide well-architected patterns and pre-built components, including innovative "AI inference templates" to help you leverage AI to create dynamic and intelligent application foundations. As an example, at launch, Application Design Center provides opinionated templates for Google Kubernetes Engine (GKE) clusters (Standard, Autopilot and NodePool) to run AI inference workloads using a variety of LLM models, as well as for enterprise-grade production clusters or single-region web app clusters.

You can also ingest and manage your existing Terraform configurations (“Bring your own Terraform”) directly from Git repositories. Once imported, you can use Application Design Center to design with your own Terraform, or in combination with Google-provided Terraform, to create standardized, opinionated infrastructure patterns for sharing and reuse across your application teams.

2. AI-powered design for rapid application designing and prototyping
Application Design Center integrates with Google's Gemini Cloud Assist Design Agent, empowering you to design actual, deployable application infrastructure application templates on Google Cloud that you can export as Terraform infrastructure-as-code.

With Gemini Cloud Assist, you can describe your application design intents using natural language. In return, Gemini interactively generates multi-product application template suggestions, complete with visual architecture diagrams and summarized benefits. You can then refine these proposals through multi-turn reasoning or by directly manipulating the architecture within the Application Design Center canvas.

Additionally, all designs that you create with Gemini are automatically observable, optimizable, and enabled for troubleshooting assistance during runtime, thanks to their tight integration with Gemini Cloud Assist.

3. A secure, sharable catalog of application templates with full lifecycle management
Platform admins can curate a collection of application templates built from Google's best-practice components. This provides developers a trusted, self-service experience from which they can quickly discover and deploy compliant applications. Tight integration with Cloud Hub transforms these governed templates into a live operational command center, complete with unified visibility into the health and deployment status of the resulting applications. This closes the critical loop between design and runtime, so that your production environments reflect your organization’s approved architectural standards.

Also, Application Design Center’s robust application template revisions serve as an immutable audit trail. It automatically detects and flags configuration drift between your intended designs and deployed applications, so that developers can remediate unauthorized changes or safely push approved configuration updates. This helps ensure continuous state consistency and compliance from Day 1 and through the subsequent evolution of your application.

4. GitOps integration automating developers’ day-to-day software design lifecycle tasks
By integrating Application Design Center into existing CI/CD workflows, platform teams empower developers to own the complete software delivery lifecycle right from their IDE. Developers can leverage compliant application and infrastructure (IaC) code using Application Design Center application templates.

Further, every infrastructure decision made through Application Design Center is committed to code, versioned, and auditable. Specifically, developers can download the application IaC template from Application Design Center and import it into their app repos (the single source of truth), clone their repo, and edit the Terraform directly in their local IDEs. Any modifications go through a Git pull request for review. Once approved, this automatically triggers the existing CI/CD setup to build, test, and deploy both app and infra changes in lockstep. This unified approach minimizes friction, enforcing "golden paths" and providing an end-to-end automated pathway from a line of code in the IDE to a fully deployed change in production.

What's new since preview

This GA launch is packed with features that users have been asking for. We’re excited to share powerful new capabilities: enterprise-grade governance and security with public APIs and gcloud CLI support; full compatibility with VPC service controls; bring your own Terraform and GitOps support for integration with your existing application patterns and automation pipelines; agentic application patterns using GKE templates (Standard, Autopilot and NodePool); and finally, a simplified onboarding experience with app-managed project support, making Application Design Center an AI-powered engine for your applications on Google Cloud.

Get started today

To help you get started, Google provides a growing library of curated Google application templates built by experts. These templates combine multiple Google Cloud products and best practices to serve common use cases, which you can configure for deployment, and view as infrastructure as code in-line. Platform teams can then create and securely share the catalogs and collaborate with teammates on designs and self-service deployment for developers. For enterprises with existing Terraform patterns and assets, Application Design Center interoperates by enabling their import and reuse within its native design and configuration experience.

Ready to experience the power of Application Design Center? You can learn more about ADC and get started building in minutes using the quickstart. You can start building your first AI-powered application template in minutes, free of cost, and quickly deploy applications with working code. For deeper insights, explore the comprehensive public documentation here. We can't wait to see how you innovate with the Application Design Center!

Replit is delivering enterprise-grade vibe coding with Google Cloud

Thu, 04 Dec 2025 17:00:00 +0000

Vibe coding has been all the rage this year. Using AI to build novel apps and create dynamic websites simply by typing conversationally into a chat interface can seem magical — yet it’s largely remained the domain of individual developers, not larger business teams.

Today, Replit and Google Cloud are expanding their strategic partnership to bring vibe coding capabilities to enterprise developers and teams.

Google Cloud was Replit’s first cloud partner, and its popular platform for AI coding continues to run on top of Google Cloud’s infrastructure, utilizing multiple Google Cloud services and integrating Google’s Gemini models through Vertex AI.

Now, Replit is extending its partnership with Google Cloud with a new, multi-year agreement that will grow its usage of infrastructure and cloud services; further integrate Google’s models into its platform; and jointly support vibe coding use cases for enterprise customers. Just last month, Replit integrated Gemini 3 into its new Design mode to positive reception. Under the terms of the expanded deal:

Google Cloud will continue to be the primary cloud provider for Replit, and services like Google Cloud Run, Google Kubernetes Engine, and BigQuery will underpin its applications and enable further scale as the company grows.
Google models, including Gemini 3, 2.5 Flash Lite, 2.5 Flash, and Imagen 4, are now supported on Replit, powering both coding and multimodal use cases — and driving significant token usage to Google Cloud.
Replit and Google Cloud will partner to help enterprise customers embrace vibe coding and help their developers be more productive — through joint go-to-market on Google Cloud Marketplace and through Google Cloud’s extensive co-sell programs.

“Our growing partnership will deliver more capabilities to Replit’s users through deeper integrations with our AI and cloud services,” Google Cloud CEO Thomas Kurian said, “and will accelerate the adoption of vibe coding in the enterprise by bringing Replit's easy-to-use AI tools, powered by Google Cloud AI, to more organizations."

Amjad Masad, co-founder and CEO of Replit, said: “Our mission is to enable the next billion software creators — from hobbyists to entrepreneurs to enterprises. Over the last few months, we have seen fantastic adoption in businesses, especially in the Fortune 1000. Today’s expanded partnership with Google will enable us to scale faster and more deeply as we integrate Google’s offerings with ours – the work is just beginning.”

You can read more about how customers are using Google’s AI models for coding use cases here, and join Replit and other customers in building with Gemini 3 via Vertex AI.

Chaos engineering on Google Cloud: Principles, practices, and getting started

Mon, 13 Oct 2025 16:00:00 +0000

As engineers, we all dream of perfectly resilient systems — ones that scale perfectly, provide a great user experience, and never ever go down. What if we told you the key to building these kinds of resilient systems isn't avoiding failures, but deliberately causing them? Welcome to the world of chaos engineering, where you stress test your systems by introducing chaos, i.e., failures, into a system under a controlled environment. In an era where downtime can cost millions and destroy reputations in minutes, the most innovative companies aren't just waiting for disasters to happen — they're causing them and learning from the resulting failures, so they can build immunity to chaos before it strikes in production.

Chaos engineering is useful for all kinds of systems, but particularly for cloud-based distributed ones. Modern architectures have evolved from monolithic to microservices-based systems, often comprising hundreds or thousands of services. These complex service dependencies introduce multiple points of failure, and it’s difficult if not impossible to predict all the possible failure modes through traditional testing methods. When these applications are deployed on the cloud, they are deployed across multiple availability zones and regions. This increases the likelihood of failure due to the highly distributed nature of cloud environments and the large number of services that coexist within them.

A common misconception is that cloud environments automatically provide application resiliency, eliminating the need for testing. Although cloud providers do offer various levels of resiliency and SLAs for their cloud products, these alone do not guarantee that your business applications are protected. If applications are not designed to be fault-tolerant or if they assume constant availability of cloud services, they will fail when a particular cloud service they depend on is not available.

In short, chaos engineering can take a team's worst "what if?" scenarios and transform them into well-rehearsed responses. Chaos engineering isn’t about breaking systems — engineering chaotically, as it were — it's about building teams that face production incidents with the calm confidence that only comes from having weathered that chaos before, albeit in controlled conditions.

Google Cloud’s Professional Service Organization (PSO) Enterprise Architecture team consults on and provides hands-on expertise on customers’ cloud transformation journeys, including application development, cloud migrations, and enterprise architecture. And when advising on designing resilient architecture for cloud environments, we routinely introduce the principles and practices of chaos engineering and Site Reliability Engineering (SRE) practices.

In this first blog post in a series, we explain the basics of chaos engineering — what it is and its core principles and elements. We then explore how chaos engineering is particularly helpful and important for teams running distributed applications in the cloud. Finally, we’ll talk about how to get started, and point you to further resources.

Understanding chaos engineering

Chaos engineering is a methodology invented by Netflix in 2010 when it created and popularized ‘Chaos Monkey’ to address the need to build more resilient and reliable systems in the face of increasing complexity in their AWS environment. Around the same time, Google introduced Disaster Resilience Testing, or DiRT, which enabled continuous and automated disaster readiness, response, and recovery of Google’s business, systems, and data. Here on Google Cloud’s PSO team, we offer various services to help customers implement DiRT as part of SRE practices. These offerings also include training on how to perform DiRT on applications and systems operating on Google Cloud. The central concept is straightforward: deliberately introduce controlled disruptions into a system to identify vulnerabilities, evaluate its resilience, and enhance its overall reliability.

As a proactive discipline, chaos engineering enables organizations to identify weaknesses in their systems before they lead to significant outages or failures, where a system includes not only the technology components but also the people and processes of an organization. By introducing controlled, real-world disruptions, chaos engineering helps test a system's robustness, recoverability, and fault tolerance. This approach allows teams to uncover potential vulnerabilities, so that systems are better equipped to handle unexpected events and continue functioning smoothly under stress.

Principles and practices of chaos engineering

Chaos engineering is guided by a set of core principles about why it should be done, while practices define what needs to be done.

Below are the principles of chaos engineering:

Build a hypothesis around steady state: Prior to initiating any disruptive actions, you need to define what "normal" looks like for your system, commonly referred to as the "steady state hypothesis."
Replicate real-world conditions: Chaos experiments should emulate realistic failure scenarios that the system might encounter in a production environment.
Run experiments in production: Chaos engineering is firmly rooted in the belief that only a production environment with real traffic and dependencies can provide an accurate picture of resiliency. This is what separates chaos engineering from traditional testing.
Automate experiments: Make resiliency testing part of a continuous ongoing process rather than a one-off test.
Determine the blast radius: Experiments should be meticulously designed to minimize adverse impacts on production systems. This requires categorizing applications and services in different tiers based on the impact the experiments can have on customers and other applications and services.

With these principles established, follow these practices when conducting a chaos engineering experiment:

Define steady state: Identifies the specific metrics (e.g., latency, throughput) that you will look at and establish a baseline for them.
Formulate a hypothesis: This is the practice of creating a single testable statement, for example, ‘By deleting this container pod, user login will not be affected’. Hypotheses are generally created by identifying customer user journeys and deriving test scenarios from them.
Use a controlled environment: While one chaos engineering principle states that experiments need to run in production, you should still start small and run your experiment in a non-production environment first, learn and adjust, and then gradually expand the scope to production environment.
Inject failures: This is the practice of causing disruption by injecting failures either directly into the system (e.g., deleting a VM, stopping a database instance) or indirectly by injecting failures in the environment (e.g. deleting a network route, adding a firewall rule).
Automate experimental execution: Automation is crucial for establishing chaos engineering as a repeatable and scalable practice. This includes using automated tools for fault injection (e.g., making it part of a CI/CD pipeline) and automated rollback mechanisms.
Derive actionable insights: The primary objective of using chaos engineering is to gain insights into system vulnerabilities, thereby enhancing resilience. This involves rigorous analysis of experimental results; identifying weaknesses and areas for improvement; and disseminating findings to relevant teams to inform subsequent experimental design and system enhancements.

In other words, chaos engineering isn't about breaking things for the sake of it, but about building more resilient systems by understanding their limitations and addressing them proactively.

Elements of chaos engineering

Here are the core elements you'll use in a chaos engineering experiment, derived from these five principles:

Experiments: A chaos experiment constitutes a deliberate, pre-planned procedure wherein faults are introduced into a system to ascertain its response.
Steady-state hypotheses: A steady-state hypothesis defines the baseline operational state, or "normal" behavior, of the system under evaluation.
Actions: An action represents a specific operation executed upon the system being experimented on.
Probes: A probe provides a mechanism for observing defined conditions within the system during experimentation.
Rollbacks: An experiment may incorporate a sequence of actions designed to reverse any modifications implemented during the experiment.

Getting started with chaos engineering

Now that you have a good understanding of chaos engineering and why to use it in your cloud environment, the next step is to try it out for yourself in your own development environment.

There are multiple chaos engineering solutions in the market; some are paid products and some are open-source frameworks. To get started quickly, we recommend that you use Chaos Toolkit as your chaos engineering framework.

Chaos Toolkit is an open-source framework written in Python that provides a modular architecture where you can plug in other libraries (also known as ‘drivers’) to extend your chaos engineering experiments. For example, there are extension libraries for Google Cloud, Kubernetes, and many other technologies. Since Chaos Toolkit is a Python-based developer tool, you can begin by configuring your Python environment. You can find a good example of a Chaos Toolkit experiment and step-by-step explanation here.

Finally, to enable Google Cloud customers and engineers to introduce chaos testing in their applications, we’ve created a series of Google Cloud-specific chaos engineering recipes. Each recipe covers a specific scenario to introduce chaos in a particular Google Cloud service. For example, one recipe covers introducing chaos in an application/service running behind a Google Cloud internal or external application load balancer; another recipe covers simulating a network outage between an application running on Cloud Run and connecting to a Cloud SQL database by leveraging another Chaos Toolkit extension named ToxiProxy.

You can find a complete collection of recipes, including step-by-step instructions, scripts, and sample code, to learn how to introduce chaos engineering in your Google Cloud environment on GitHub. Then, stay tuned for subsequent posts, where we’ll talk about chaos engineering techniques, such as how to introduce faults into your Google Cloud environment.

Inside the AI-powered assistant helping doctors work faster and better at Seattle Children’s Hospital

Thu, 18 Sep 2025 16:00:00 +0000

Though its name may suggest otherwise, Seattle Children’s is the largest pediatric healthcare system in the world.

While its main campus is in its namesake city, Seattle Children’s also encompasses 47 satellite hospitals across Alaska, Montana, Idaho, and Washington, and patients come from as far away as Hawaii for treatment. For more than 100 years, Seattle Children’s has helped kids across the Western U.S. get healthy and stay healthy, regardless of the ability to pay.

With so much ground to cover and diverse patient populations to treat, Seattle Children's has always looked to new technologies to bring improved, consistent care to its patients and providers. Generative AI is now the latest advance in their medical toolkit.

It started roughly two decades ago, when Seattle Children’s created its pediatric clinical pathways, a set of standardized protocols designed to help clinicians make quicker and more reliable decisions to address dozens of medical conditions. Such pathways were becoming commonplace across medicine, and Seattle Children’s had developed some of the first for children’s unique medical needs.

Innovative as these were, they still required clinicians to thumb through indexes and long binders of information to find what they needed for a given ailment. And in healthcare, it’s often the case that every second counts.

Seattle Children’s was already working with Google Cloud on a number of projects, and as we began to explore the potential for generative AI to make the work of our clinicians easier, the clinical pathways seemed like an obvious place to start. Using Vertex AI and Gemini, we were able to quickly develop our Pathways Assistant, which took training from the clinical pathways documentation and supercharged it with not just searchability but conversationality.

Instead of flipping pages, we’d flipped the script on how quickly and reliably clinicians could find the lifesaving information they needed.

aside_block: <ListValue: [StructValue([('title', '2025 ROI of AI in Healthcare and Life Sciences'), ('body', <wagtail.rich_text.RichText object at 0x7fe0b8162b50>), ('btn_text', 'Read now'), ('href', 'https://cloud.google.com/resources/content/roi-of-ai-healthcare-life-sciences'), ('image', None)])]>

The pathways to improved healthcare run through Gemini

“Clinical pathways” are end-to-end treatment protocols for a specific condition or illness. Seattle Children’s pediatric clinical pathways are widely respected and used by hospitals around the globe, providing information on everything from diagnostic criteria to testing protocols to medication recommendations.

Previously, these clinical pathways were documented exclusively in PDFs — hundreds of thousands of pages of them. Performing a traditional search of their contents for the answers clinicians needed delayed their ability to provide treatment in an environment where minutes or even seconds can be critical.

Google Cloud engineers worked with Seattle Children’s informatics physicians, who straddle the worlds of healthcare and technology, to create Pathway Assistant. The new multimodal AI chatbot that responds to spoken or written natural-language queries using the information in those PDFs.

After processing a question, Pathway Assistant searches each PDF’s metadata, which contains semi-structured data in JSON format that’s been extracted from the PDFs by Gemini and curated by clinicians. It then selects the most relevant PDFs, parses the information — including any complex flowcharts, diagrams, and illustrations embedded in them — and answers the clinician’s question in just a few seconds.

Interactive information-finding for accurate decision-making

Pathway Assistant becomes more accurate with use. Healthcare providers can “discuss” clinical pathways with the chatbot, which, instead of answering a question, poses questions of its own if it needs clarification, going back and forth until it’s confident it can answer accurately. The chatbot always displays the specific sections of each PDF that was the source for formulating its answers, helping clinicians confirm the veracity of responses.

The interface also includes a way for users to provide feedback about the accuracy and appropriateness of the chatbot’s analysis and answers. The feedback is then logged in a BigQuery table for future forensic analysis — both by clinicians, who can query the information using natural language, and by the built-in Gemini models, which processes the feedback and summarizes what clinicians found confusing or how to improve the accuracy of future answers.

This reflexive capability enables Pathway Assistant to update the PDFs based on clinicians’ feedback if the inaccuracy stemmed from the PDF’s content. Clinicians are also finding that the metadata is becoming more accurate and requiring less curation. Pathway Assistant even corrects typos in the documentation automatically. And as new clinical pathways are developed, PDFs containing the latest information are added to the PDF library.

This growing collection is housed securely in Google Cloud Storage, and the bigger it gets, the more useful it becomes — which wasn't always the case. Whereas an expanding paper-based collection contained more information, it was also more material to wade through, which is especially challenging in emergency medical situations. Pathway Assistant almost entirely relieves this burden, synthesizing and delivering the most complete information at any time in a matter of seconds.

Ultimately, Pathway Assistant is not a decision-making tool but rather an information-finding tool. Research into critical, evidence-based guidelines that used to take hours now takes minutes.

This speed and effectiveness helps clinicians make the right decisions more quickly at the point of care, drastically reducing research time and improving patient safety and outcomes. Ultimately, clinicians can spend more time with more patients, not with more PDFs.

Ask any physician, they’ll tell you that’s what the best medical technology enables them to do — focus on the patient, not paperwork.

How Rent the Runway supercharges developer speed and insights with Cloud SQL

Fri, 12 Sep 2025 16:00:00 +0000

Editor’s note: Rent the Runway is redefining how consumers engage with fashion, offering on-demand access to designer clothing through a unique blend of e-commerce and reverse logistics. As customer expectations around speed, personalization, and reliability continue to rise, Rent the Runway turned to Google Cloud’s fully managed database services to modernize its data infrastructure. By migrating from a complex, self-managed MySQL environment to Cloud SQL, the company saved more than $180,000 annually in operational overhead. With Cloud SQL now supporting real-time inventory decisions, Rent the Runway has built a scalable foundation for its next chapter of growth and innovation.

Closet in the cloud, clutter in the stack

Rent the Runway gives customers a “Closet in the Cloud” — on-demand access to designer clothing without the need for ownership. We offer both a subscription-based model and one-time rentals, providing flexibility to serve a broad range of needs and lifestyles. We like to say Rent the Runway runs on two things: fashion and logistics. When a customer rents one of our items, it kicks off a chain of events most online retailers never have to think about.

That complex e-commerce and reverse logistics model involves cleaning, repairing, and restocking garments between uses. To address both our operational complexity and the high expectations of our customers, we’re investing heavily in building modern, data-driven services that support every touchpoint.

Out: Manual legacy processes

Our legacy database setup couldn’t keep up with our goals. We were running self-managed MySQL, and the environment had grown… let’s say organically. Disaster recovery relied on custom scripts. Monitoring was patchy. Performance tuning and scaling were manual, time-consuming, and error-prone.

Supporting it all required a dedicated DBA team and a third-party vendor providing 24/7 coverage — an expensive and brittle arrangement. Engineers didn’t have access to query performance metrics, which meant they were often flying blind during development and testing. Even small changes could feel risky.

All that friction added up. As our engineering teams moved faster, the database dragged behind. We needed something with more scale and visibility, something a lot less hands-on.

It’s called Cloud SQL, look it up

When we started looking for alternatives, we weren’t trying to modernize our database: We were trying to modernize how our teams worked with data. We wanted to reduce operational load, yes, but also to give our engineers more control and fewer blockers when building new services.

Cloud SQL checked all the boxes. It gave us the benefits of a managed service — automated backups, simplified disaster recovery, no more patching – while preserving compatibility with the MySQL stack our systems already relied on.

But the real win was the developer experience. With built-in query insights and tight integration with Google Cloud, Cloud SQL made it easier for engineers to own what they built. It fit perfectly with where we were headed: faster cycles; infrastructure as code; and a platform that let us scale up and out, team by team.

aside_block: <ListValue: [StructValue([('title', 'Build smarter with Google Cloud databases!'), ('body', <wagtail.rich_text.RichText object at 0x7fe09aef03d0>), ('btn_text', ''), ('href', ''), ('image', None)])]>

Measure twice and cut over once

We knew the migration had to be smooth because our platform didn’t exactly have quiet hours. (Fashion never sleeps, apparently.) So we approached it as a full-scale engineering project, with phases for everything from testing to dry runs.

While we planned to use Google’s Database Migration Service across the board, one database required a manual workaround. That turned out to be a good thing, though: It pushed us to document and validate every step. We spent weeks running Cloud SQL in lower environments, stress-testing setups, refining our rollout plan, and simulating rollback scenarios. When it came time to cut over, the entire process fit within a tight three-hour window. One issue did come up – a configuration flag that affected performance — but Google’s support team jumped in fast, and we were able to fix it on the spot.

Overall, minimal downtime and no surprises. Our kind of boring.

Less queues and more runway

One of the biggest changes we’ve noticed post-migration was how our teams worked. With our old setup, making a schema change meant routing through DBAs, coordinating with external support, then holding your breath during rollout. Now, engineers can breathe a little easier and own those changes end to end. Cloud SQL gives our teams access to IAM-controlled environments where they can safely test and deploy. That has let us move toward real CI/CD for database changes — no bottlenecks or surprises.

With better visibility and guardrails, our teams are shipping higher-quality code and catching issues earlier in the lifecycle. Meanwhile, our DBAs can focus on strategic initiatives — things like automation and platform-wide improvements — rather than being stuck in a ticket queue.

Our most cost-effective look yet

Within weeks of moving to Cloud SQL, we were able to offboard our third-party MySQL support vendor, cutting over $180,000 in annual costs. Even better, our internal DBA team got time back to work on higher-value initiatives instead of handling every performance issue and schema request.

Cloud SQL also gave us a clearer picture of how our systems were running. Engineers started identifying slow queries and fixing them proactively, something that used to be reactive and time-consuming. With tuning and observability included, we optimized instance sizing and reduced infrastructure spend, all without compromising performance. And with regional DR configurations now baked in, we’ve simplified our disaster recovery setup while improving resilience.

In: Cloud SQL

We’re building toward a platform where engineering teams can move fast without trading off safety or quality. That means more automation, more ownership, and fewer handoffs. With Cloud SQL, we’re aiming for a world where schema updates are rolled out as seamlessly as application code.

This shift is technical, but it’s also cultural. And it’s a big part of how we’ll continue to scale the business and support our expansion into AI. The foundation is there. Now, we’re just dressing it up.

Learn more:

Discover how Cloud SQL can transform your business! Start a free trial today!
Download this IDC report to learn how migrating to Cloud SQL can lower costs, boost agility, and speed up deployments.
Learn how Ford and Yahoo gained high performance and cut costs by modernizing with Cloud SQL.

Driving for the Horizon: New Android Automotive solution on cloud offers faster builds

Mon, 08 Sep 2025 06:00:00 +0000

The automotive industry is in the midst of a profound transformation, accelerating towards an era of software-defined vehicles (SDVs). This shift, however, presents significant challenges for manufacturers and suppliers alike. Their priority is making great vehicles, not great software, though the latter now contributes — and is increasingly a necessity — to achieve the former. These OEMs must find ways to bring greater efficiency and quality to their software delivery and establish new collaboration models, among other hurdles to achieving their visions for SDVs.

To help meet this moment, we’ve created Horizon, a new open-source software factory for platform development with Android Automotive OS — and beyond. With Horizon, we aim to support the software transformation of the automotive industry and tackle its most pressing challenges by providing a standardized development toolchain so OEMs can generate value by focussing on building products and experiences.

In early deployments at a half-dozen automotive partners, we’ve already seen between 10x to 50x faster feedback for developers, leading to high-frequency releases and higher build quality. In this post we will outline how Horizon helps overcome the key impediments to automotive software transformation.

The Roadblocks to Innovation in Automotive Software Development

Today, traditional automotive manufacturers (OEMs) often approach software development from a hardware-centric perspective that lacks agility and oftentimes struggles to scale. This approach makes software lifecycle support burdensome and is often accompanied by inconsistent and unreliable tools, slowing down development.

OEMs face exploding development costs, quality issues and slow innovation, making it difficult to keep pace with new market entrants and the increasing demand for advanced features. Furthermore, most customers expect frequent, high-quality over-the-air (OTA) software updates similar to what they receive on other devices such as on their smartphones, forcing most OEMs to mirror the consumer electronics experience.

But a car is not a television or refrigerator or even a rolling computer, as many now describe them. Vehicles are made up of many separate, highly complex systems, which typically require the integration of numerous components from multiple suppliers who often provide "closed box" solutions. Even as vehicles have become more connected, and dependent on these connective systems for everything from basic to advanced operations, the vehicle platform has actually become harder, not easier, to integrate and innovate with.

We knew there had to be a better way to keep up with the pace necessary to provide a great customer experience.

Introducing HORIZON: A Collaborative Path Forward

To tackle these pressing industry challenges, Google and Accenture have initiated Horizon. It is an open-source reference development platform designed to transform the automotive industry into a software-driven innovation market.

Our vision for Horizon is enabling automakers and OEMs to greatly accelerate their time to market and increase the agility of their teams while significantly reducing development costs. Horizon provides a holistic platform for the future of automotive software, enabling OEMs to invest more in innovation rather than just integration.

Key Capabilities Driving Software Excellence

Horizon offers a comprehensive suite of capabilities, establishing a developer-centric, cloud-powered, and easy-to-adopt open industry standard for embedded software.

1. Software-First Development with AAOS

Horizon champions a virtual-first approach to product design, deeply integrating with Android Automotive OS (AAOS) to empower software-led development cycles. This involves the effective use of the vehicle hardware abstraction layer (VHAL), virtio, and high-fidelity cloud-based virtual devices like Cuttlefish, which can scale to thousands of instances on demand. This approach allows for scalable automated software regression tests, elastic direct developer testing strategies, and can be seen as the initial step towards creating a complete digital twin of the vehicle.

2. Streamlined Code-Build-Test Pipeline

Horizon aims to introduce a standard for the entire software development lifecycle:

Code: It supports flexible and configurable code management using Gerrit, with the option to use GerritForge managed service via our Google Cloud Marketplace for productive deployments. With Gemini Code Assist, integrated in Cloud Workstations, you can supercharge development by leveraging code completion, bug identification, and test generation, while also aiding in explaining Android APIs.
Build: The platform features a scaled build process that leverages intelligent cloud usage and dynamic scaling. Key to this is the caching for AAOS platform builds based on warmed-up environments and the integration of the optimized Android Build File System (ABFS), which can reduce build times by more than 95% and allow full builds from scratch in one to two minutes with up to 100% cache hits. Horizon supports a wide variety of build targets, including Android 14 and 15, Cuttlefish, AVD, Raspberry Pi devices, and the Google Pixel Tablet. Build environments are containerized, ensuring reproducibility.
Test: Horizon enables scalable testing in Google Cloud with Android’s Compatibility Test Suite (CTS), utilizing Cuttlefish for virtualized runtime environments. Remote access to multiple physical build farms is facilitated by MTK Connect, which allows secure, low-latency interaction with hardware via a web browser, eliminating the need for hardware to be shipped to developers.

3. Cloud-Powered Infrastructure

Built on Google Cloud, Horizon ensures scalability and reliability. Deployment is simplified through tools like Terraform, GitOps and Helm charts, offering a plug-and-play toolchain and allowing for tracking the deployment of tools and applications to Kubernetes.

Unlocking Value for Auto OEMs and the Broader Industry

The Horizon reference platform delivers significant benefits for Auto OEMs:

Reduced costs: Horizon offers a reduction in hardware-related development costs and an overall decrease in rising development expenses.
Faster time to market: By accelerating development and enabling faster innovation cycles, Horizon helps OEMs reduce their time to market and feature cycle time.
Increased quality and productivity: The platform enables stable quality and boosts team productivity by providing standardized toolsets and fostering more effective team collaboration.
Enhanced customer experience: By enabling faster, more frequent and higher-quality builds, OEMs can change the way they develop vehicle software, thus offering enhanced customer experiences and unlocking new revenue streams through software-driven services.
Strategic focus: Horizon underpins the belief that efficient software development platforms should not be a point of differentiation for OEMs; instead, their innovation should be focused on the product itself. This allows OEMs to devote more time and resources to software development with greater quality, efficiency, and flexibility.
Robust ecosystem: To ensure scalable, secure, and future-ready deployments across diverse vehicle platforms, Horizon aims to foster collaboration between Google, integration partners, and platform adopters. While advancing the reference platform capabilities, Horizon also allows for tailored integration and compatibility with vehicle hardware, legacy systems and compliance standards.

The Horizon ecosystem

It’s been said that the best software is the one you don’t notice, so seamless and flawless is its functioning. This is especially true when it comes to the software-defined vehicle, where the focus should be on the road and the joy of the trip.

This is why we believe the platforms enabling efficient software development shouldn’t be differentiating for automakers — their vehicles should be. Like a solid set of tires or a good sound system, software is now essential, but it’s not the product itself. That is the full package put together by the combination of design, engineering, development, and production.

Because software development is now such an integral part of that process, we believe it should be an enabler, not a hindrance, for automakers. To that end, the Google Cloud, Android, and Accenture teams have continuously aimed to simplify access and the use of relevant toolchain components. The integration of OpenBSW and the Android Build File System (ABFS) are just the latest waypoints in a journey that started with GerritForge as providing a managed Gerrit offering, and continuing with additional partners in upcoming releases.

Please, join us on this journey. We invite you to become a part of the community to receive early insights, provide feedback, and actively participate in shaping the future direction of Horizon. You can also explore our open-source releases on Github to evaluate and customize the Horizon platform by deploying it in your Google Cloud environment and running reference workloads.

Horizon is a new dawn for the future of automotive software, though we can only get there together, through open collaboration and cloud-powered innovation.

^{A special thanks to a village of Googlers and Accenture who delivered this,}^{Mike Annau, Ulrich Gersch, Steve Basra, Taylor Santiago, Haamed Gheibi, James Brook, Ta’id Holmes, Sebastian Kunze, Philip Chen, Alistair Delva, Sam Lin, Femi Akinde, Casey Flynn, Milan Wiezorek, Marcel Gotza, Ram Krishnamoorthy, Achim Ramesohl, Olive Power,}^{Christoph Horn, Liam Friel, Stefan Beer, Colm Murphy, Robert Colbert, Sarah Kern, Wojciech Kowalski, Wojciech Kobryn, Dave M. Smith, Konstantin Weber, Claudine Laukant, Lisa Unterhauser}

^—

^{Opening image created using Imagen 4 with the prompt: Generate a blog post header image for the following blog post, illustrating the concept of a software-defined vehicle <insert the first six paragraphs>.}