Won 2nd place at HopHacks 2025 for Best Data-Driven Hack. Sponsored by Marshall Wace

An interactive visualization platform that maps tweets with location data and provides AI-powered trend explanations including sentiment analysis through Gemini API.

• Real-time tweet mapping with geographic visualization
• Topic-based tweet filtering and exploration (Top 25 trends are shown, top 50 can be obtained via API access)
• AI-powered topic explanations using Google Gemini 2.0 flash
• Location-based tweet clustering as well as location autocomplete

• React + Typescript + Tailwindcss with deck.gl for map visualization
• Rust with SpacetimeDB for storage and live feed
• Google Gemini API for topic analysis
• Apify for tweet collection API
• Flask hosting a small FLAN model along with geocoding API for location guessing

• GET /api/trends
  • Returns an array of today's trending topics on twitter for the last hour
• GET /api/tweets
  • Returns all tweets collected on server
• GET /api/tweets/:topic
  • Returns a specific array of tweets tagged with a topic
• GET /api/flattened
  • Returns all tweets in a topic, longitude, latitude, text, author format
• GET /api/flattened/:topic
  • Returns all tweets in a topic, longitude, latitude, text, author format for a specific topic

A special thanks to to HopHacks 2025, Sponsors at SpacetimeDB, and Google Gemini

Copyright (c) 2025 Ruslan Akmyradov, Daniel Yi