Microsoft Builds the Agentic Web: Transforming Development in 2025 and Beyond
At Build 2025, Microsoft unveiled a bold vision for the future of the open, agentic web at scale. Moving away from vertically integrated application stacks, Microsoft is establishing a comprehensive platform designed to enable an open, scalable agentic web. This platform aims to unlock unprecedented opportunities for developers across every layer of the technology stack.
Empowering Developers with Enhanced Tools
Central to this transformation is Visual Studio, recognized as one of the most powerful IDEs for .NET and C++. Microsoft announced significant enhancements, including:
- Full support for .NET 10
- Live preview capabilities during design time
- Enhanced Git tooling
- A new debugger optimized for cross-platform applications
In parallel, Visual Studio Code marked its 100th release, introducing improved multi-window support and streamlined staging directly within the editor, reinforcing its status as a premier developer tool.
Microsoft’s ecosystem further thrives with GitHub, which continues to gain momentum in the enterprise sector. With an increased emphasis on trust, security, compliance, auditability, and data residency, GitHub remains the preferred platform for developers building modern applications.
AI as a Core Development Pillar
Artificial Intelligence has become integral to Microsoft’s approach to coding assistance. GitHub Copilot, the AI-powered pair programmer, has been open-sourced within VS Code, embedding AI capabilities at the heart of this widely used open-source editor.
Beyond coding assistance, Copilot now supports application modernization by upgrading legacy frameworks, such as Java 8 to Java 21 and .NET 6 to 9, and automating the migration of on-premises applications to cloud environments.
Microsoft also introduced the Azure SRE Agent, an autonomous tool for site reliability engineering that automatically triages, diagnoses, and mitigates issues, logging incident reports as GitHub issues. Repair tasks can be assigned directly to Copilot, which operates as a full coding peer—autonomously fixing bugs, adding features, and maintaining codebases.
Integrated AI Experiences: Chat, Search, Notebooks, and Agents
Microsoft has unified AI-powered chat, search, notebooks, content creation, and agents into a seamless, intuitive interface described as the UI for AI. This system integrates both web and enterprise data, enabling powerful knowledge discovery across platforms such as Confluence, Google Drive, Jira, ServiceNow, and Microsoft 365.
The Notebooks feature allows creation of heterogeneous data collections, combining chats, documents, emails, and more, with capabilities to generate audio reviews and podcasts. The Create function supports conversion of PowerPoint slides into explainer videos and instant image generation.
Among specialized AI agents, the Researcher agent synthesizes information from web and enterprise sources using advanced reasoning, while the Analyst agent processes raw data from multiple sources to deliver insights, forecasts, and visualizations.
All agents integrate with Microsoft Teams and Copilot, enabling interaction via chat or meetings, assignment of tasks, and initiation of workflows. The Teams AI library simplifies building multi-agent applications with support for multi-channel processing and advanced memory features powered by Azure Search.
Enterprise Customization with Copilot Tuning
Recognizing the uniqueness of each organization, Microsoft offers Copilot Tuning—allowing AI models to be fine-tuned on company-specific data, workflows, and linguistic styles. This results in AI assistants that understand corporate tone and expertise, deployed securely with proper permission controls.
Complementing this, the Model Router for Azure OpenAI automatically selects the optimal AI model for each task, enabling true multi-model applications. Microsoft also announced the arrival of Grok from XAI on Azure, expanding the range of AI models available.
Azure AI Foundry: Declarative Agents and Lifecycle Management
The Azure AI Foundry and Agent Foundry platforms enable developers to build complex, multi-agent orchestrations with minimal code. The Foundry agent service, now generally available, supports deployment to containerized apps, Azure Functions, and hybrid environments using Azure Arc.
Developers can fine-tune or post-train AI models within Foundry and integrate them into Copilot Studio for workflow automation or agent creation. A notable example is Stanford’s healthcare agent orchestrator, now accessible to the wider developer community.
Advanced observability tools in Foundry allow comprehensive monitoring of AI deployments, tracking impact, quality, safety, and cost from a centralized dashboard. With integration into Entra ID, agents gain individual identities, permissions, and access policies. Partnerships with ServiceNow and Workday enable automated provisioning and management of agents through Entra.
For data governance, Microsoft Purview integrates with Foundry, ensuring end-to-end protection. Security is further enhanced by Microsoft Defender, which protects AI agents against threats like credential theft and wallet abuse.
Extending AI to Edge and Native Windows Support
Microsoft extends AI capabilities beyond the cloud with Foundry Local, providing high-performance AI runtime, agent-as-a-service, and local app development tools on Windows and Mac platforms.
The Windows AI Foundry platform supports development across CPUs, GPUs, NPUs, and cloud environments, enabling applications to run seamlessly on diverse hardware. It taps into a rich catalog of pre-optimized open-source models for local execution.
Windows now includes native support for Multi-Channel Processing (MCP) with built-in MCP servers for filesystems, settings, and windowing, along with a secure MCP registry ensuring vetted performance and security.
Additionally, the Windows Subsystem for Linux (WSL) has been made fully open source, marking a decade since its initial launch and enhancing developer flexibility.
Data Integration and AI-Driven Applications
Microsoft introduced NL Web, an open-source framework that enables easy transformation of existing websites or APIs into agentic applications.
Cosmos DB is integrated directly into Foundry, allowing AI agents to store conversational histories and support retrieval-augmented generation (RAG) use cases. Azure DataBricks connections facilitate linking AI workflows to enterprise data, while Postgres supports direct language model responses within SQL queries.
These developments unify structured and semi-structured data—including text, images, and audio—within the Fabricdata platform, preparing data estates for AI workloads.
The Digital Twin Builder, embedded in Fabric, enables rapid, no-code creation of digital twins to map physical assets and systems. AI-powered Shortcut Transformations provide pre-built ETL processes such as audio-to-text and sentiment analysis, integrated into Fabric workflows.
Azure also hosts the world’s largest GPU-based supercomputer, designed to scale AI workloads for broad accessibility.
Accelerating Scientific Research with AI
Microsoft’s Discovery platform leverages Foundry to accelerate scientific workflows. It interprets complex scientific knowledge from public and proprietary data sources, enabling advanced reasoning and research automation. Industries such as biopharma are already benefiting from Discovery’s capabilities to enhance R&D productivity.
Microsoft is reshaping the software development landscape through a comprehensive, system-wide platform approach to AI and the agentic web. By delivering open-source AI tools, enterprise-grade customization, multi-agent orchestration, edge computing support, and domain-specific AI applications, Microsoft empowers developers and organizations to create smarter, more scalable, and responsive software. This vision not only drives innovation but also expands possibilities across industries, marking a new era in application development and AI-driven transformation.
Originally published in the VerbionPulse newsletter on LinkedIn: Microsoft Builds the Agentic Web: Transforming Development in 2025 and Beyond.
Comments