Microsoft Reactor Sydney

Want to build smarter AI solutions—fast? This session is designed for developers ready to move beyond theory and understand how to create real-world AI applications. Explore Microsoft Foundry and the AI Toolkit (AITK) in VS Code to compare the latest multimodal and reasoning models from the Model Catalog. Learn how to augment models for business scenarios using prompt and context engineering. See how to prototype an agent with AITK Agent Builder and equip it with the right tools via MCP for secure, enterprise-ready workflows.

Technologies Used

• AI toolkit for Visual Studio Code
• GitHub models
• Microsoft Foundry Models
• MCP - Model Context Protocol

🧠 What you’ll learn:

• Pick the right models for your scenario
  Explore and compare the latest multimodal and reasoning models in the Microsoft Foundry Model Catalog to find the best models for your project.
• Boost accuracy with context engineering
  Learn how to refine prompts and add business context for precise, actionable results.
• Prototype intelligent agents fast
  Use AITK Agent Builder and MCP to quickly build agents that act on your business logic

Explore the future of multi-agent collaboration with the Microsoft Agent Framework, built by the Semantic Kernel and AutoGen engineering teams. Learn how to design A2A-compatible agents using SWE Agents (GitHub Copilot coding agent + Codex with Azure OpenAI models) and implement MCP tools for secure orchestration.

This session demonstrates how to build, deploy, and scale enterprise-grade agentic systems, leveraging pre-release capabilities for advanced workflows.

🧠 What You’ll Learn:

• Use Microsoft Agent Framework to unify agents and cut integration complexity.
• Accelerate coding with SWE Agents and achieve rapid prototyping.
• Leverage MCP to keep sensitive data safe while orchestrating agents for enterprise compliance.