{"type":"video","version":"1.0","html":"<iframe src=\"https://www.loom.com/embed/0694157509ca49ba9cf51f637f6c571c\" frameborder=\"0\" width=\"1670\" height=\"1252\" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>","height":1252,"width":1670,"provider_name":"Loom","provider_url":"https://www.loom.com","thumbnail_height":1252,"thumbnail_width":1670,"thumbnail_url":"https://cdn.loom.com/sessions/thumbnails/0694157509ca49ba9cf51f637f6c571c-e37968fa60062246.gif","duration":198.416,"title":"Kaizen - Next Evolution of Code Reviews","description":"This Loom introduces Kaizen, a tool for detecting drifting architecture and product decisions by turning repo code into a single source of truth for review. Kaizen builds an intent layer that product and engineering leaders can approve, plus engineer review of the decisions, and it also guides coding agents using the same knowledge. The speaker demonstrates Kaizen mapping a marketing app built from scratch in Webflow, extracting PRDs like goals, non-goals, and user stories, and generating tech specs including an append-only constitution state machine and an access model deterministic miner. Kaizen is also dogfooded on the Paisan repo and tested on another public home audio distribution project, extracting its domain model, purpose, and supporting documentation from the code alone."}