Keynote — Noise and Signal
summary
Renaud Chevalier examines where real value lies amid GenAI/LLM/GPU/assistant/agent hype by asking whether AI agents can replace a product team. He reports on a hands-on experiment showing an AI agent delivering a product-team task in under 10 minutes at ~7/10 quality for under €2, discusses implications for teams and organizations, and argues that agile practitioners must shift from optimization to daring responses to this new reality.
AI-written summary · about
chapters (22)
- 0:00Keynote introduction
- 0:51Generative AI noise
- 2:15Personal upskilling approach
- 3:18From research to demo
- 4:01N8N presentation and tech choices
- 4:42First N8N agent
- 7:10Automated user stories
- 10:23Jira integration demo
- 12:29Scrum team automation
- 15:57Workflow cost and performance
- 23:46Prompt evaluation and industrialization
- 25:52Vibe coding intro
- 28:36AI-assisted dev tools
- 31:23Chains vs agents
- 33:17Agent autonomy and memory
- 36:04MCP protocol and connected tools
- 37:57Autonomous agent demo
- 40:28RAG and contextual enrichment
- 43:18Sub-agents and parallel orchestration
- 47:38Replacing Scrum team with agents
- 50:20Pomai and product ops model
- 52:07AI reflections and emotions
related talks
10mKeynote: The Last Mile Problem: Why AI W...
01h 01mKeynote Panel: Future of AI agents
32mKeynote: Will the Real Autonomous Agent Please Stand Up
49mKeynote: Operations and AI with Jen Krieger
33mKeynote: Boring Wins in AI Agents 🔥 Hard-Earned Lessons
24mKeynote - The AI Playbook For Companies
57mWhen AI Agent Teams Work (And When They Just Create More Work)
38mKeynote: Thriving in the Age of AI