Operate - Autopilot for engineering ops
Problem: I spent 4 years at Microsoft building Xbox and Bing at scale. Then 9 years building software platforms for 30+ companies backed by Sequoia, Techstars, and Google. Across every engagement, the same pattern kept showing up — engineering teams losing 30-40% of capacity to production investigations. Support tickets, error triage, alert storms — all manual, all unplanned, all pulling the best engineers away from building.
Solution: I am building Operate (operatex.dev) to fix this. It's a self-hosted AI agent that investigates production incidents end-to-end — pulls your logs, queries your DB read-only, reads git history, finds root cause, opens a draft PR. Nothing merges without an engineer's sign-off.
Working with early design partners who feel this pain strongly. If your engineering team is losing hours to unplanned investigations, I'd like to talk.
