Migrating 400 Messy Tables in 2 Days Using an Agent Fleet
Deepfeet-09 · reddit · 2026-07-31
A team facing the migration of 400 messy legacy database tables post-acquisition ditched manual ETL scripts in favor of an autonomous agent fleet to inspect schemas, infer mappings, and run tests.
Key Lessons:
- Decomposition: A single massive prompt failed constantly on nested JSON and hallucinated column types. Success came only after breaking tasks into tiny, single-purpose agents.
- Adversarial Loops: Never let the executing agent verify its own work. The team built a reviewer agent to intentionally try breaking queries, sending raw SQL errors back to the mapper for fixing—saving production from breaking.
- State Management: Managing memory across retries was painful until they integrated Lyzr to isolate memory and prevent unmasked data leaks.
The approach reduced the human review queue to under 4%, compressing a quarter-long task into just two days.
More from coding & agent
- Connected steps are not a workflow: nobody designs the handoff — usehive · 2026-10-16
- Making Blender animations with Opus, Astra and Argon: how you prompt still matters — chongdashu · 2026-10-16
- Running PolyStrata + Cline coding agent on a $0-fancy laptop at ~90 tok/s — rpwoerk · 2026-10-16
- How do you attribute agent gains: model training or infrastructure? — jrxna · 2026-10-16
- Feed Claude your books, blogs and top engineers' code for claude.md — 'it now writes better code than me' — blaizedsouza · 2026-10-16
- Open-source /label plugin auto-names and colors Claude Code sessions in one step — hermanschutte · 2026-10-16