Inspiration
When you put down your information on Hinge, Tinder, Bumble, etc, you are 'faking' a bit of your personality to look more attractive. TasteMatch analyzes your real YouTube data to create authentic dating profiles that show who you actually are, in return, will find you a person which you'll instantly connect with!
What it does
TasteMatch uses AI to analyze your YouTube subscriptions and liked videos to generate comprehensive personality profiles, then matches you with compatible people based on deep personality analysis rather than superficial swiping.
How we built it
Built with Next.js, Supabase, Gemini, and OpenAI for personality analysis. We integrated Google OAuth with YouTube API access to fetch user data and created a modern dating interface with AI-powered compatibility scoring.
Challenges we ran into
Getting accurate personality insights from limited YouTube data was challenging. We also struggled with creating a matching/compatibility algorithm that balances personality traits, interests, and dating preferences effectively, so we instantly used Gemini (and OpenAI as a fallback) to do all the matching/compatibility for us.
Accomplishments that we're proud of
We created a fully functional dating platform with AI personality analysis that actually works. The compatibility scoring system provides detailed explanations for why users match, making connections more meaningful.
What we learned
AI can extract surprisingly accurate personality insights from digital behavior patterns. We also learned how to build a complex full-stack application with real-time matching and database management.
What's next for TasteMatch
Adding video profile introductions, location-based matching, group date suggestions, and expanding to other social media platforms for even more comprehensive personality analysis.
Built With
- claude
- css3
- cursor
- gemini
- nextauth.js
- nextjs
- openai
- sql
- supabase
- tailwindcss
- tsx
- typescript

Log in or sign up for Devpost to join the conversation.