tabletalk

• 

  logo

• 

  interacting w/ AI

• 

  tabletalk w/ other interested ppl in La Contenta Les NYC

Inspiration

Restaurant discovery apps focus on browsing and ratings, but they don’t help you actually go. I wanted to turn discovery into action. Instead of “want to try,” make it “let’s go.” Young 20-somethings are wanting to meet each other in cities and make friends, what better way than through food.

What it does

Users swipe through personalized recommendations, join “Tables” (groups of people interested in the same restaurant), and create plans together in a chat. The app leverages Yelp AI and Groq APIs to provide personalized restaurant insights and a conversational assistant for each restaurant.

How we built it

React Native with Expo for cross-platform mobile development, Yelp APIs for restaurant data and AI insights, Groq AI for conversational features, TypeScript for type safety, and AsyncStorage for offline data storage.

Challenges we ran into

Yelp and Groq APIs have rate limits --> implemented 24-hour caching. No backend infrastructure for hackathon submission --> built client-side architecture ready for eventual backend integration, used mock data for demo.

Accomplishments that we're proud of

Implemented Yelp AI API (not mock templates) for personalized content, built a caching system that improves performance and reduces costs, smooth UX for users, core features are functional offline as well, restaurant assistant for each restaurant.

What we learned

Offline-first architecture patterns, API integration best practices, React Native optimizations.

What's next for tabletalk

Backend infrastructure for real social features, database, push notifications for plan updates, WebSocket RTC, user auth. Eventually, calendar app integrations.

Built With

• expo.io
• groq
• node.js
• react-native
• typescript
• yelp-ai-api