Multi-Agent Systems & Autonomous Engineering (Featuring Eno Reyes, CTO of Factory.AI) скачать в хорошем качестве

Multi-Agent Systems & Autonomous Engineering (Featuring Eno Reyes, CTO of Factory.AI)

Did you check the Stanford Terminal-Bench lately? If you look at the leaderboard, the #1 Coding Agent isn't the one everyone is talking about on Twitter. It's Factory.AI – a system that is currently outperforming raw GPT-5.2 and Claude Opus 4.5 on complex terminal tasks. So, how did they build it? We are hosting Eno Reyes, Co-founder & CTO of Factory.AI, to open up the hood and show us how production-grade Multi-Agent systems really work. This isn't a product pitch. It's an architectural deep dive. We will explore the engineering required to move from simple "Copilots" to fully autonomous Droids: The Benchmark: Understanding the architecture that hit 64.9% on Terminal-Bench (and why standard LLMs fail there). -️ Orchestration in Production: How to wrangle 20+ agents working in parallel without chaos (The "Gastown" model). - Solving Amnesia: Implementing persistent memory ("Beads") for long-running engineering tasks. Who is this for? Engineers who want to understand how autonomous agents are built, orchestrated, and deployed in production.