Keynote: Next-Gen AI Applications via Multi-Agent Conversation

on YouTube ↗10m

" summary

AutoGen is presented as an open source multi-agent conversation framework that composes specialized LLM agents (including a user proxy agent, an assistant agent and a group chat manager) to plan, act with tools and validate results; examples include a two‑agent code+execution loop that fixes dependency errors, a critic agent that improves visualizations, and a book generation run where 12 messages took about 3 minutes. Repo metrics: 18,000 stars, ~150 contributors, 11,000 members on Discord.

AI-written summary · about

" related talks

• 34mBuilding Multi-Agent Generative-AI Applications with AutoGenColombo et al. · PyData London 2024
• 45mBuild Agentic AI Apps with the Autogen frameworkNitya Narasimhan, Jingya Chen, Eric Zhu · Microsoft Build 2024
• 25mFrontiers of Agentic AIChi Wang · Open Agent Summit 2025
• 01h 01mKeynote Panel: Future of AI agentsBhanu Prakash Manjappasetty Masagali, Ganapathi Pulipaka, Suresh Veeragoni, Dileep Kumar Pandiya, Yogi Sikri · Global Artificial Intelligence Conference 2024
• 30mKeynote: Fireside ChatEmil Eifrem, Andrew Ng · NODES 2025
• 26mAI powered automation & multi-agent orchestrationShawn Henry · AI Dev Days 2025
• 06mKeynote: Building Agents at ScaleTiffany Souterre · Open Source Summit Europe 2025
• 20mGenAI agents - Evolution of automationArunansu Pattanayak · Global Artificial Intelligence Conference 2024