Consumer App · AI
Date Vibes
An AI date ideas platform that plans the night out — personalized suggestions built from real local venues and tuned to mood, budget, weather, and location.
The problem
Planning a date night is a research project. You open maps, then reviews, then a reservation app, then the weather, and forty minutes later you go back to the same restaurant you always go to. The information needed to plan a good night out exists — it is just scattered across five apps that do not talk to each other.
What we built
A planner that answers the actual question: what should we do tonight? Couples pick a vibe — romantic, adventurous, chill, impress, late night — plus a budget and a time, and the app assembles a complete multi-stop date from real venues near them: somewhere to eat, something to do, somewhere to close the night. Each stop comes with the reason it was picked, so the suggestion is arguable rather than arbitrary.
Recommendations that sharpen
Preferences and venues are embedded and matched semantically with pgvector, so the match is on the character of a place rather than a category label. Every swipe feeds back into the profile, and the recommendations narrow toward what a specific couple actually likes. Weather and location are inputs, not afterthoughts — a rainy Tuesday and a clear Saturday produce different nights.
How it is built
A React Native iOS app on the App Store, a React marketing site, and a NestJS API over PostgreSQL with pgvector handling the embedding search. One backend serves both surfaces, so the app and the site never disagree about what a venue is.
What we delivered
- AI recommendations grounded in real local venues, not generic lists
- pgvector semantic matching that sharpens with every swipe
- React Native iOS app shipped alongside the marketing site
More work
Let's talk about your project
Tell me what you're building and I'll get back to you within one business day — no sales runaround.
Get in touch