Skip to main content Get to know Azure Microsoft as Customer Zero View all products (200+) Microsoft Foundry Azure Copilot GitHub Copilot Azure Kubernetes Service (AKS) Azure Cosmos DB Azure Database for PostgreSQL Azure Arc Microsoft Fabric Linux virtual machines in Azure Foundry Models Foundry Agent Service Foundry IQ Foundry Tools Foundry Control Plane Observability in Foundry Control Plane Azure OpenAI in Foundry Models Azure Speech in Foundry Tools Azure Machine Learning View all databases Azure Cosmos DB Azure DocumentDB Azure SQL Azure Database for PostgreSQL Azure Managed Redis Microsoft Fabric Azure Databricks Linux virtual machines in Azure Windows Server on Azure Azure Functions Azure Virtual Machine Scale Sets Azure API Management Azure Container Apps Azure Kubernetes Service (AKS) Azure Kubernetes Fleet Manager Azure Container Registry Azure Red Hat OpenShift Azure Container Instances Azure Container Storage Azure Arc Azure Local Microsoft Defender for Cloud Azure Monitor Microsoft Sentinel Azure Catalyst View all solutions (40+) Cloud solutions for small and medium businesses Cloud migration and modernization center Data analytics for AI Azure Databases AI apps and agents Microsoft Marketplace Microsoft Sovereign Cloud AI apps and agents Responsible AI with Azure AI Infrastructure Data analytics for AI Machine learning operations (MLOps) Low-code application development on Azure Integration Services Serverless computing DevOps Migration and modernization center .NET apps migration Databases on Azure Linux on Azure Oracle on Azure SAP on the Microsoft Cloud Adaptive cloud High-performance computing (HPC) Infrastructure as a service (IaaS) Resiliency Azure Essentials Frontier Accelerate for Azure FinOps on Azure Microsoft Marketplace Azure pricing overview Create an Azure account Free Azure services Flexible purchase options Pricing calculator FinOps on Azure Maximize ROI from AI Azure savings plans Azure reservations Azure Hybrid Benefit Virtual Machines Azure SQL Microsoft Foundry Microsoft Fabric Azure Kubernetes Service (AKS) Microsoft Defender for Cloud View more Software Development Companies Microsoft Marketplace Find a partner Resources for Azure partners Get started with Azure Customer stories Analyst reports, white papers, and e-books Videos Learn more about cloud computing Documentation Explore Azure portal Developer resources Quickstart templates Resources for startups Developer community Students Azure for partners Blog Events and Webinars Learn Support Contact Sales Get started with Azure Sign in

As AI continues to reshape how every business builds, operates, and competes, it is rapidly emerging as the defining workload in the cloud. This new class of agents, more demanding, complex, and dynamic than anything before them, is rewriting the rules of cloud infrastructure. Reasoning across multi-agent systems, drawing on rich context, acting in real time, and running continuously, they ask more of every layer of the platform beneath them. This week, we’re introducing innovations across Microsoft Azure that set a new standard for how the cloud powers AI: expanded AI core capabilities in Azure to build what’s next, and Azure Catalyst, an intelligent agent-led solution to modernize the apps, data, and infrastructure that already run the business.

For years, we have been reengineering every layer of Azure for the unique demands of AI and proving it on the world’s most advanced AI workloads, from leading frontier models to our own GitHub Copilot, Microsoft Teams, and Microsoft 365. That work has forged Azure into a single, co-engineered system, with models, silicon, and cloud infrastructure tuned as one and value compounding at every layer. This week, we take that even further.

New capabilities across isolated execution, agent networking, context management, and inference efficiency deliver faster, more secure agents with better unit economics. Every innovation rests on an open foundation, giving customers the freedom to run open models and open systems everywhere they operate. There is a wave of new AI applications scaling in production, and years of existing apps and data are moving to the cloud, becoming AI-ready. Azure was built for this moment.

Of everything we’re announcing this week, three areas matter most:

  1. Building the next generation of AI applications. Silicon, compute, storage, and networking work as one system across Azure for inference, training, and agents. A new connected path from Microsoft Foundry to Azure Kubernetes Service (AKS) gives you choice in how and where you run AI.
  2. Modernizing apps and data with agents. Azure Catalyst uses AI agents to modernize applications, data, and infrastructure in a fraction of the time and cost of traditional methods, turning decades of technical debt into the fastest path to AI.
  3. Extending AI wherever you operate. Azure Local, flex nodes for AKS and Microsoft Foundry extend AI across cloud and distributed environments, governed through Azure Arc. Windows Server 2028, now in public preview, and Windows PCs, bring AI closer to applications, data, and users, providing the choice to meet workload-specific data, performance, regulatory, and cost needs.

Build and run the next generation of AI applications

Microsoft Foundry and AKS are at the center of how you build and run AI from Foundry’s managed experience, to the deep customization of AKS across models, runtimes, performance, security, and infrastructure.

The connected path between them runs from building AI applications and agents to running them at scale. Open-weight models and agents move seamlessly from Foundry to AKS clusters, deployed from cloud to edge, with consistent infrastructure, security policies, and governance.

Underneath these experiences, AI applications are challenging infrastructure in new ways: writing and executing code autonomously, connecting dynamically to multiple agents, models, tools, and data, and carrying context across long-running tasks. We’re advancing the infrastructure to meet each of these demands:

  • Isolation and confidentiality. With the Agent Sandbox extension for AKS, agent-generated code runs in secure, managed sandboxes with virtual machine (VM)-level isolation for every session, pre-warmed pools for fast startup, and a serverless option with no infrastructure to manage. Azure confidential computing extends protection to sensitive data and workloads while in use.
  • Agent networking and security. With Agent Grid, agents securely connect to other agents, models, tools, data, and Azure services through authenticated, encrypted connections. Identity-driven policies and default-deny network isolation control how agents communicate and what resources they can access.
  • Context and state. Long-context, long-running agents reuse processed prompt context instead of recomputing it on the GPU, so long tasks run faster and cost less. That cached context stays durable and controlled, by you, in your Azure tenant.

And we’re making Azure itself more intelligent, building in new workload-aware capabilities that combine deep context for what each AI application requires with what only the platform delivers: GPU and VM performance, topology, and real-time availability across the fleet. AKS AI Manager translates what each model needs into the right infrastructure, placing each deployment where the fleet has the GPU capacity to serve it best. The result is a cloud that makes every AI workload perform better, with faster responses, better economics, and optimal placement decisions.

And across all of this, we’re extending our commitment to open-source, giving you transparency and access to the best innovation shaped by the community. Kubernetes, vLLM, and Ray, together with flex nodes for AKS in preview, bring greater flexibility from development through production, including supported GPU environments beyond Azure. You have the freedom to build on the best model, use the right framework, and run where your business needs it.

We’re already seeing what’s possible with these innovations. Rockwell Automation is using AI in its factories to help engineers diagnose equipment issues and investigate product failures. As these applications move into production, Rockwell wants to build on what works and bring those approaches to other facilities:

Running our factory-level AI applications on Azure gives us one foundation for different AI workloads in production. When something works in one plant, we can adapt it for the next instead of starting over.

—Justin Litherland, Vice President, Enterprise Program Management, IT, Rockwell Automation

Within our own teams, Microsoft 365 engineering built COSMIC, a shared platform on AKS, to help them ship faster with built-in security, compliance, and reliability. Making infrastructure easier to run matters for developers. COSMIC handles infrastructure centrally so developers can focus on customer experiences rather than Kubernetes operations. What we learn as our own customer continues to shape future Azure capabilities for you.

To go deeper at Microsoft Ignite, check out these sessions:

  • BRK420: Deploy and run Microsoft Foundry models and agents on Kubernetes
  • BRK448: Run secure agent loops in production with isolated sandboxes at scale
  • BRK429: The evolution of Azure infrastructure for AI models, apps, and agents

Modernize applications and data with Azure Catalyst

The need to modernize is universal. But with developers stretched thin, custom code and deep running dependencies, progress often stalls and the backlog keeps growing. Decades of technical debt adds security exposure, cost and complexity, while disconnected tools create friction from assessment to execution.

Agents in Azure Catalyst change that, analyzing legacy code, mapping dependencies, and planning and executing with precision. Work that once took months or years can now be done in weeks or even days, even with limited resources and skills, making it possible to modernize systems that were previously out of reach.

The results are real: 1.4 million applications modernized, with businesses like Levi Strauss modernizing its enterprise systems and deploying more than 1,000 agents across them.

Azure Catalyst modernizes end to end—across applications, data, and infrastructure in one solution—delivered in the tools hundreds of millions of developers and IT professionals already use. Here’s what sets it apart:

One connected experience

Azure Catalyst connects IT teams and developers across an end-to-end process from discovery to design and execution. New capabilities add rich estate insights, AI-powered application discovery, modernization strategy and cost comparisons, and AI-generated migration waves. Execution spans applications, containers, storage, networking, and databases, and builds on existing coverage for Java and .NET modernization and migration support for VMware, Hyper-V, and physical servers.

Tailored to your business

Governance rulebooks and bring-your-own-model support let you set the standards, guardrails, and models agents follow, so plans fit your architecture, priorities, and security requirements. Reusable runbooks and skills make what works repeatable, so every project meets your exact standards.

Secure from the start

Outdated applications and data increase security exposure. Azure Catalyst works with Microsoft Security to find, prioritize, and eliminate vulnerabilities end to end. Together they scan and rank risk, prioritize actions by severity and exposure, and flag vulnerable code and databases for modernization. Through GitHub Copilot you can quickly remediate vulnerabilities as applications are modernized, fixing them once and shipping as both modern and hardened.

Ready for AI

Azure Catalyst works in GitHub Copilot to help teams understand the business logic and dependencies of complex applications and upgrade those apps to run on Azure’s managed services. These upgrades make it easier to connect apps to AI workflows. New previews support modernizing applications and databases together and migrating from BizTalk to Azure Logic Apps. And because Azure Catalyst is built for Azure, it optimizes performance, security, and economics along the way.

Microsoft is customer zero for agent-led modernization, putting the same agentic tooling in Azure Catalyst across our estate. Microsoft Quantum and Xbox achieved 80% faster app modernization, while Intune upgraded a key library 75% faster than anticipated. GitHub ported the GitHub Copilot agentic harness—800,000 lines of production code—from TypeScript to Rust. By using these technologies ourselves, we learn firsthand what it takes to modernize complex, production-scale systems, from targeted fixes to large-scale code transformation, and apply those lessons to help you modernize your own environment.

For a full breakdown, read the Azure Catalyst announcement blog. To go deeper at Microsoft Ignite, check out these sessions:

  • BRK412: Rethink modernization end to end with agents
  • BRK409: Modernize faster with AI-powered discovery, planning, and execution
  • BRK410: Accelerating Java, .NET, and database modernization with agents

Run AI across cloud, edge, and devices

Across cloud, enterprise infrastructure, edge, physical systems, and personal devices, we’re giving you the power to run AI where your business needs it. Azure Local and flex nodes for AKS extend cloud infrastructure and Kubernetes compute into distributed environments, while extending Microsoft Foundry across these environments. Azure Arc provides consistent infrastructure management and governance end to end. New at Microsoft Ignite, Windows Server 2028 enables AI inference alongside business-critical applications and data, while Windows PCs bring intelligence directly to the devices where people work through local processing. Together, you compose the right infrastructure for each AI workload to best meet data, performance, regulatory, and cost requirements.

For a full breakdown, read the Windows Server 2028 blog. To go deeper at Microsoft Ignite, check out these sessions:

  • BRK418: What’s new and what’s next for Windows Server
  • BRK442: Build sovereign AI with Azure Local and Foundry Local

Innovate on an open foundation for what comes next

The architecture decisions you make today will carry your business for the next decade. Our job is to ensure the systems you depend on now can evolve into an AI-powered future, and that everything you build next runs on an open, co-engineered platform with choice over your models and where your workloads run. The real test of everything we’re introducing is how these solutions perform in your environment. We’ll keep listening and building alongside you.

This week, I’m looking forward to Mark Russinovich, Chief Technology Officer (CTO), Deputy Chief Information Security Officer (dCISO), and Technical Fellow, Microsoft Azure and Pavan Davuluri, Executive Vice President, Windows and Devices, discussing why Azure and Windows are the foundation for mission-critical AI, from datacenters to the edge to a billion devices, at INV643: AI infrastructure from cloud to edge. We’d love to see you there.

Microsoft Azure

Azure unifies infrastructure, data, applications, and AI on a trusted cloud platform to accelerate innovation and scale AI with confidence.

WE ARE MICROSOFT

Explore Microsoft Foundry

The future of AI starts here. Envision your next great AI app with the latest technologies. Get started with Azure.