Inspiration
Hackathons are all about judging projects, but we started wondering: if judges are judging us… who’s judging the judges? 👀 In the spirit of Hack&Roll, we wanted to flip the script, have some fun, and build something memorable. So we decided to hold the judges accountable :) (lovingly, and with a healthy dose of humour).
What it does
Uno-Reverse secretly evaluates the judges while they evaluate you. We built a deliberately terrible decoy project to provoke reactions from judges, while our real system ran quietly in the background. Uno-Reverse: Captures judges’ facial expressions and attention levels using computer vision Transcribes judges’ verbal feedback using OpenAI’s Whisper Analyses both visual and verbal cues to infer engagement, confusion, and emotional responses Based on this data, it generates a humorous “performance review” for the judges — giving them feedback on how they could give better feedback.
How we built it
With equal parts chaos and caffeine. Behind the scenes, Uno-Reverse was built using: Python for computer vision and emotion analysis OpenAI Whisper for speech-to-text transcription TypeScript for orchestration and integration Alongside that, we engineered a fully functional intentionally awful decoy project to distract the judges — complete with cursed UI and painful UX.
What we learned
What we learned Sometimes the best hackathon ideas come from not taking things too seriously Building a “joke” project still requires solid engineering Platform-agnostic tools can have real value beyond the meme While Uno-Reverse started as a joke, we realised it has real potential as a background analytics tool for presentations or multi-round judging environments — helping teams understand when audiences lose interest, get confused, or disengage. Also, programming is still about the friends we made and the snacks we ate along the way.
Built With
- human
- javascript
- openai-whisper
- python
- react
- typescript
Log in or sign up for Devpost to join the conversation.