Stealth software factory reports patterns from 12,000 agentic code reviews over six months

Watson Labs reported that its stealth software factory Gymwasp processed 372 issues over six months using agentic workflows with no human code review, routing humans instead to plan validation and product decisions. The factory deployed seven parallel reviewer agents per change, each with narrow briefs and exclusion lists to avoid duplicate findings, drawing on 12,000 agentic code reviews. Watson Labs said most teams lack measurement systems for agentic review effectiveness beyond deployment speed or issue volume.

Stealth software factory reports patterns from 12,000 agentic code reviews over six months According to Watson Labs, a stealth software factory Gymwasp using agentic workflows processed 372 issues over six months with no human code review, instead routing humans to plan validation and product decisions. The factory deployed seven parallel reviewer agents per change, each with narrow briefs and exclusion lists to avoid duplicate findings. Watson Labs reports that most teams lack measurement systems for agentic review effectiveness beyond deployment speed or issue volume. Topics Sources - Press Read article https://blog.watson-labs.co.uk/6-learnings-from-12000-agentic-code-reviews/ Go deeper This intelligence is sourced automatically from public sources across the web and synthesised by the Prefactor AI pipeline. Stories are reviewed before publication.