Foundry Friday AMA · Jul 11, 2025 · AI & Developer Experiences

[nitya]

AMA on AI Developer Experiences

This is part of the #ModelMondays series where we put the spotlight on a new model-related topic each week.

This post was generated with AI help and human revision & review. To learn more about our motivation and workflows, please refer to this document in our website

🌟🌟 See #54 for the full Foundry Fridays AMA schedule 🌟🌟

---

Event Details

Ready to streamline your AI development workflow? This session highlights the latest tools and extensions for building, evaluating, and deploying generative AI apps. Leo Yao will showcase the AI Toolkit and Azure AI Foundry extensions for Visual Studio Code, sharing tips for model selection, agent development, and integrating AI into your projects. Whether you’re new to AI or a seasoned developer, discover how to boost productivity and bring your ideas to life with Microsoft’s developer ecosystem.

• 1️⃣ | Register for the Friday AMA - 1:30pm ET
• 2️⃣ | Watch the Monday Livestream - 1:30pm ET
• 3️⃣ | Learn more About Model Mondays - Season 1 Recaps + Season 2 Schedule

Related Resources

• What is the AI Toolkit for Visual Studio Code?
• Work with the Azure AI Foundry for Visual Studio Code extension
• Work with Azure AI Foundry Agent Service in Visual Studio Code

---

[nitya]

Model Mondays S2:E04 — AI Developer Experiences: AI Toolkit & Azure AI Foundry Extensions

This post was generated with AI help and human revision & review.
To learn more about our motivation and workflows, please refer to this document in our website

---

Browse The Slides

👉🏽 EXPLORE THESE 3 LINKS FIRST:

1. AI Toolkit for Visual Studio Code - Download the extension!
2. Build An Agent Series - Watch the video series
3. Agent Support Series - Read the blog posts

---

Watch The Replay

Abstract

This episode explores the latest advancements in AI developer tooling at Microsoft, focusing on the AI Toolkit and Azure AI Foundry extensions for Visual Studio Code. The session features a live demo, practical guidance for building and deploying AI agents, and highlights from Microsoft’s research and product teams.

Highlights

• Microsoft’s new medical AI model and Project Aurora for atmospheric science were spotlighted as cutting-edge research.
• GitHub Copilot Agent and Azure S Agent are making developer and operations workflows more autonomous and efficient.
• Project Emily introduces an autonomous agent for machine learning engineering tasks, streamlining ML workflows.
• The AI Toolkit for VS Code now supports model catalog browsing, local and cloud deployment, and agent building with integrated tools.
• Live demo showed how to deploy models, build agents, connect to MCP servers, and generate code for programmatic use.

Spotlight

The spotlight segment featured Leo Yao, Product Manager at Microsoft, who demonstrated the end-to-end workflow for building generative AI agents in Visual Studio Code using the AI Toolkit and Azure AI Foundry extensions. Leo walked through model catalog browsing, deploying models to Azure, and using the agent builder to create both local and cloud-based agents. The demo included connecting agents to Model Context Protocol (MCP) servers, configuring tools like Bing search and code interpreter, and generating code for further development. The session emphasized the flexibility of bringing your own models, integrating local or remote resources, and the ease of transitioning from visual design to code-first workflows. Leo also addressed audience questions about MCP server management, multi-model support, and code generation for programmatic agent deployment.

Takeaways

• Microsoft’s AI Toolkit and Azure AI Foundry extensions streamline the process of building, deploying, and managing AI agents in VS Code.
• Developers can leverage a rich model catalog, deploy models locally or in the cloud, and integrate tools like MCP servers and Bing search.
• The extensions support both visual and code-first workflows, making them accessible to a wide range of developers.
• New research and product features, such as Project Amelie and GitHub Copilot Agent, are pushing the boundaries of AI-driven productivity.
• Community engagement and feedback are encouraged to drive further improvements in these tools.

Related Resources

• AI Toolkit for Visual Studio Code
• Get started with AI Toolkit for Visual Studio Code
• Azure AI Foundry for Visual Studio Code extension
• Work with Azure AI Foundry Agent Service in Visual Studio Code
• How to use the Model Context Protocol (MCP) tool

[nitya]

Episode 4 AMA Recap: AI Developer Experiences

This post was generated with AI help and human revision & review.
To learn more about our motivation and workflows, please refer to this document in our website

Attendees

Main Speakers:

• Leo Yao - Microsoft
• April Gittens - Microsoft
• Nitya Narasimhan, PhD - Microsoft
• Lee Stott - Microsoft

Community Participants:

• 15+ active community members from various backgrounds and experience levels
• Strong participation from India, USA, South Africa, and other global locations

Locations

• New York, USA
• Bangalore, India
• Dallas, USA
• South Africa
• Global community participation from various time zones

Core Topics Discussed

1. AI Toolkit for Visual Studio Code

• Extension installation and setup
• Model catalog browsing and GitHub Models integration
• Agent Builder functionality for creating AI agents
• Local vs. cloud model deployment strategies

2. Azure AI Foundry Integration

• VS Code extension capabilities
• Model deployment to Azure AI Foundry
• Agent creation and deployment workflows
• Integration with existing Azure services

3. Model Context Protocol (MCP)

• MCP server integration within AI Toolkit
• Custom MCP server creation
• Community experiences with MCP implementations
• GitHub Copilot integration with MCP

4. Evaluation and Testing

• Built-in evaluators (F1 score, relevance, coherence, similarity)
• Custom evaluator creation
• RAG-enhanced LLM evaluation capabilities
• Agent quality measurement strategies

5. GitHub Models

• Rate limits for free and enterprise tiers
• Integration with AI Toolkit
• Prototyping capabilities and limitations

Key Takeaways

1. Unified Development Experience: AI Toolkit provides a seamless experience from model discovery to agent deployment, all within VS Code, eliminating the need to switch between multiple tools.

2. Flexible Model Access: Developers can choose between GitHub Models for free prototyping, local models for offline development, or Azure AI Foundry for production deployments.

3. MCP Integration is Mature: The toolkit supports both consuming existing MCP servers and creating custom ones, with comprehensive documentation and hands-on labs available.

4. Evaluation is Built-in: The platform includes comprehensive evaluation tools for measuring agent quality, with both built-in and custom evaluator support.

5. Community-Driven Learning: Regular series like "Agent Support" blog posts and video tutorials provide ongoing learning resources for developers at all levels.

Featured Links

AI Toolkit

• AI Toolkit Extension - Main VS Code extension
• AI Toolkit Video Series - Tutorial series
• AI Toolkit Documentation - Quick access to documentation
• AI Toolkit Evaluation System - Support for Evaluations
• Custom Evaluator Creation - Create custom evaluators
• Agent Builder Documentation - Build agents and prompts in AI Toolkit
• Agent Support Blog Series - Weekly developer advice column
• AI Toolkit July Update - Latest features and improvements

Other Topics

• Let's Learn MCP Events - Beginner's guide to MCP
• MCP for Beginners - Complete curriculum with hands-on labs
• Use MCP servers in Visual Studio - MCP support in Visual Studio 2022
• GitHub Models Marketplace - Free model access for prototyping
• GitHub Models Documentation - Rate limits and usage guide

---

• Next Session: Register for Fine-Tuning Session
• Discussion Forum: Continue the conversation
• Watch The Replay: - Model Mondays Playlist