Evolving Intelligence: Designing Autonomous AI Agents
summary
Practical steps for building autonomous agents with .NET and Azure AI Foundry: code examples and a runnable repo that instantiate clients, use the engine service for prototyping, wire an agent thread or custom memory store, and attach tools as functions (weather and GitHub demos shown) via MCP and OpenAI function spec. The talk demos agent workflows, middleware, scratchpad looping, and tracing for evaluation, compares models (GPT5 Grok4 Anthropic) and notes pre release packages, costs, and what actually worked in demos.
AI-written summary · about
chapters (16)
- 0:00Welcome
- 1:112025 the year of AI agents
- 2:14Defining autonomous agents
- 4:00Azure AI Foundry
- 7:33Semantic Kernel to Agent Framework
- 10:04Agent in code
- 12:15Agent memory and context
- 15:23Tools and function calling
- 17:16MCP for external services
- 19:44Middleware for execution
- 21:56Agent metacognition
- 27:20Agent workflows
- 28:27Choosing agent models
- 33:22Models in Azure AI Foundry
- 35:28Challenges in agentic systems
- 38:00Closing
related talks
58mAdvanced Autonomous AI Agents: Memory, Collaboration and Evolution in Action
36mBuilding AI Agents in .NET with Azure AI Foundry
25mBuilding AI Agents
56mBuilding AI Agents and Multi-Agent Systems with Azure AI
01h 54mBuilding Code First AI Agents with Azure AI Agent Service
01h 30mCreating Intuitive AI Agents for PowerShell..
05h 17mBuilding AI Applications
01h 08mFrom Model Integration to Agents: Building Intelligent .NET Applications