Code Analysis: AI software docs from interviews, SQL and screens

Hi Product Hunt! I’m Alexandre, the maker of Code Analysis. I built it while analysing a system being implemented. Interviews helped me understand the work, but I needed a way to turn that knowledge into use cases. The product grew from there into SQL documentation, screen-based documents and user manuals. Code Analysis keeps the sources and the work together: • Organize interviews, SQL and screenshots in a project. • Draft diagrams, data dictionaries and manuals with AI. • Choose an eligible model and review the credit estimate. • Refine documents in contextual chat and revisit version history. The library includes 26 configurable diagram types and notations. PlantUML source is editable and exportable; public external diagram rendering is currently disabled. AI output needs professional review. The gallery uses fictional examples, and the video includes AI-generated scenes. Neither is a live speed benchmark. Plans are individual and priced in BRL. AI and metered file transfers share one CodeCredits balance. There is no public free trial. Access is currently by invitation while we prepare the public opening. If you analyse, implement or document software, which part of turning scattered knowledge into useful documentation is hardest for you? I’d love your feedback.