3 ingredients for building reliable enterprise agents
summary
Three concrete levers raise expected value for enterprise agents: pick high‑value tasks or shift to longer, ambient workflows so agents do substantial work; increase reliability by making parts more deterministic workflows and adding observability and eval (example: Lang Smith) to shrink error bars; and reduce the cost of failures with reversible actions (commits/PRs, diffs like Replet) plus human-in-the-loop approval patterns and time‑travel rollbacks.
AI-written summary · about
chapters (6)
- 0:00Building reliable agents
- 2:57Increasing agent value
- 5:10Improving reliability
- 9:46Reducing cost of errors
- 13:12Scaling with ambient agents
- 17:38Concrete example and QA
related talks
25mFrom Chaos to Control: Building Enterprise-Grade LLM Agents That Actually Work
20mThe Three Ways AI Agents Fail. Here's How to Fix Yours.
30mArchitecting Integrity: Responsible Software Evolution in the AI Era
29mBuilding Trust in AI Agents
33mThe Autonomous Enterprise
16mMaking Codebases Agent Ready
14mOrchestrating Agentic Workflows with Governance and Trust
46mPanel Discussion: The Agentic Software Factory That Enterprises Actually Need