Microsoft Build 2025 Keynote: Everything Revealed, in 14 Minutes

Summary

In 2025, Microsoft is transforming the development landscape with groundbreaking advancements, shifting from vertically integrated apps to a platform-driven, open agentic web. Key updates include Visual Studio's enhancements, the open-sourcing of GitHub Copilot, autonomous agents for site reliability engineering, and the introduction of multi-model capabilities with Grock from XAI. Developers can now build, deploy, and manage agents across Teams and GitHub, enhancing workflows and collaboration. New tools like Foundry and Microsoft Discovery are revolutionizing R&D, allowing faster innovation in industries like healthcare. Microsoft is empowering developers to create scalable, intelligent systems that will define the future of technology.

Takeaways

• 😀 Microsoft is advancing the open, scalable agentic web, moving from vertically integrated apps to a platform that empowers developers at all levels.
• 😀 Visual Studio 2025 introduces new features including .NET 10 support, live preview at design time, and cross-platform debugger improvements.
• 😀 GitHub is doubling down on security, trust, compliance, and data residency, with GitHub Copilot playing a key role in AI-driven development.
• 😀 GitHub Copilot is evolving, now able to assist with framework upgrades, app modernization, and autonomous issue triaging in site reliability engineering.
• 😀 New autonomous agents are being introduced to automate tasks like bug fixes, feature development, and code maintenance directly within GitHub Copilot.
• 😀 Microsoft Teams now integrates AI agents, enabling chat, search, and task management within workflows, and providing a multiplayer agent development experience.
• 😀 AI capabilities in Teams and Copilot are expanding with features like researcher agents that synthesize web and enterprise data for deep insights.
• 😀 New enterprise-grade agents can be built using company-specific data, offering fine-tuned models that understand a company’s tone, workflows, and expertise.
• 😀 Microsoft announces the integration of Grock from XAI into Azure, enabling multi-model capabilities for more efficient AI model provisioning and orchestration.
• 😀 The Foundry agent service is now generally available, supporting multi-agent orchestration for complex workflows and providing integration with cloud and hybrid environments.

Q & A

What is the primary goal of the 2025 vision mentioned in the transcript?

The primary goal is to build an open, scalable, and agentic web, transitioning from a few vertically integrated apps to a platform that empowers developers to create at scale.

What new features are introduced for Visual Studio and VS Code in the transcript?

For Visual Studio, new features include support for .NET 10, live previews, Git tooling improvements, and a cross-platform debugger. For VS Code, the 100th release includes better multi-window support and improved Git integration.

How has GitHub evolved in terms of its integration with AI?

GitHub has integrated AI into its tools, particularly through GitHub Copilot, which is now open-sourced and embedded in VS Code. Copilot now functions as an autonomous coding assistant capable of upgrading frameworks, migrating apps, and handling various programming tasks.

What is the function of the new autonomous agent for site reliability engineering (SRE)?

The new autonomous agent for SRE automatically triages and mitigates issues, then logs incident reports as GitHub issues with actionable repair items, which can be assigned to GitHub Copilot for further action.

How do AI-powered agents help with chat, notebooks, and data management?

AI agents provide enhanced functionalities such as synthesizing data across web and enterprise sources, creating heterogeneous collections of data in notebooks, and automating workflows by interacting through chat, pages, and other documents.

What is Copilot Tuning and how does it benefit enterprises?

Copilot Tuning allows Copilot to learn a company's unique tone, language, and expertise, helping create custom AI models tailored to a company’s specific needs. This enables more efficient workflows and automation.

What new capabilities are introduced with multi-model support in the Azure Foundry platform?

Azure Foundry now supports multi-model provisioning, allowing developers to use multiple AI models for different tasks without manual model selection. This includes the integration of Grock from XAI for enhanced model provisioning.

How does Microsoft support AI model management and security in production environments?

Microsoft provides tools for managing AI in production through Foundry, which offers observability features to track impact, quality, and cost. Additionally, agents are protected with security measures like Defender integration to safeguard against threats.

What is the significance of Foundry Local and how does it improve edge computing?

Foundry Local allows AI agents and models to run locally on devices, such as Windows and Mac, with high-performance runtimes. This capability supports edge computing by enabling AI applications to be run directly on local hardware.

What is the role of Microsoft Discovery and how does it impact R&D?

Microsoft Discovery is designed for research and development workflows, integrating AI agents specialized in R&D to accelerate scientific discoveries by synthesizing both public domain and private data sources. It aims to enhance the productivity of R&D teams.