An intelligent, AI-powered retrospective assistant for agile teams using the 4Ls framework.
Many agile teams run retrospectives that feel routine or poorly structured. Without a clear format, it's hard to capture what went well, what was learned, what was missing, and what the team wants next.
Action items often get lost, and there's no simple way to turn discussion into a shareable report or to tie outcomes to practices like Disciplined Agile.
I designed and built YodaAI as an AI-powered assistant that guides sprint reviews using the 4Ls framework: Liked, Learned, Lacked, and Longed For. The tool leads the facilitator and the team through each category so that feedback is structured and comparable across sprints.
Under the hood, it uses natural language processing to group similar comments into themes, so the team can see patterns (e.g. "communication," "tooling," "scope") without manually tagging everything.
A voting mechanism lets the team prioritise which themes matter most, and the app suggests Disciplined Agile–aligned practices that could address those priorities. After the session, YodaAI generates a PDF report summarising the 4Ls, the main themes, and the chosen action items, so that the retrospective is documented and easy to share.
Teams get consistent, structured retrospectives with automatic theme detection, prioritisation, and report generation. Action items are captured and traceable, and recommendations are aligned with established agile practices.