Maersk Engineering: Translating SOPs and 100k Corrections for Agents
AI Engineer · youtube · 2026-08-30
Maersk's Dmitry Buykin shares engineering insights on deploying AI Agents in global shipping. The core challenge is translating human-readable screenshot SOPs into executable agent workflows (preconditions, decisions, backend calls).
Key Points:
- System is the Loop: The refining loop around the agent is the real system. Maersk spent 9 months making over 100,000 corrections; the procedure corpus outweighs runtime code 20:1.
- Accuracy is Earned: Accuracy comes from corrections, not upfront design. Heat maps turn traces into priorities.
- Context Matters: A single shipping step implies different things in different countries.
- Discovery vs. Production: Discovery needs agent freedom; production needs a cage to prevent dumb mistakes.
More from coding & agent
- Void v1.1.11 open sourced: one-command build and deploy to Cloudflare — cnakazawa · 2026-10-15
- Proof Launches x401, an Open Protocol Letting AI Agents Prove Who They Act For — csuwildcat · 2026-10-15
- Dev pushes back on "AI coding means insecure apps": humans always shipped bugs — dreamwieber · 2026-10-15
- Eight non-negotiable principles for building AI that actually learns from experience — tokenbender · 2026-10-15
- Animated teardown of XBOW's Chrome full-chain exploit that won Google's first $250K VRP bounty — moyix · 2026-10-15
- pstack's durable core: 10-20 line skills that distill agent best practices into principles — HankYeomans · 2026-10-15