Agentic Dev Partner Academy Keynote: GenAI and Intelligent Agents
16 June 2025
Yesterday I gave the opening keynote at the Agentic Dev Partner Academy, a 1-hour session on the latest innovations in GenAI and intelligent agents, and how they can transform developer workflows, automate complex tasks, and improve productivity. Here is the video recording.
Abstract
Patrick Chanezon, VP of Developer Relations at Microsoft, will guide you through the latest innovations in GenAI and intelligent agents. Discover how agents can transform your workflows, automate complex tasks, and improve productivity.
Key themes
The talk covered several themes I’ve been exploring over the past year:
- The evolution of GenAI: from large language models to multi-modal, multi-step reasoning agents that can plan, use tools, and collaborate with humans in complex workflows.
- Intelligent Agents in practice: how agents go beyond simple chat interactions to orchestrate multi-step tasks, call APIs, browse the web, write and execute code, and collaborate with other agents.
- The Open Agentic Web: protocols like MCP (Model Context Protocol) and A2A (Agent-to-Agent) that enable interoperability between agents and services.
- Transforming developer workflows: how AI Coding Agents (GitHub Copilot, Claude Code, Gemini CLI) are changing the role of developers from individual contributors to managers of AI agent teams.
- Building agents with Microsoft tools: Azure AI Foundry, Copilot Studio, and the GitHub Copilot SDK as platforms for building and deploying intelligent agents.
- Productivity gains: practical patterns for delegating complex tasks to agents, evaluating their output, and integrating them into existing team workflows.
The developer as agent boss
A central message of the talk: as AI agents become more capable, every developer needs to learn new skills around context engineering, specification-driven development, and composing teams of specialized agents. What got you here won’t get you there — the transition from hands-on-keyboard coding to managing agents requires a mindset shift similar to moving from an IC to a management role.
Resources
If you want to dive deeper into some of the topics covered:
- Reinventing Developer Relations in the age of AI Agents — my detailed post on how AI agents are transforming DevRel
- Achieve more with AI Agents — slides and video from my earlier talk on AI Agents
- MCP for Beginners — getting started with Model Context Protocol
- Awesome GitHub Copilot Customizations — community curated list of Copilot customizations
- AGENTS.md — open format for providing context to coding agents
- GitHub Agent HQ — the open ecosystem for all coding agents on GitHub