Project Inspiration
beacon was built to make it easier for people in urgent situations to find nearby essential services like food, shelters, showers, clinics, legal help, and more. We wanted to solve the problem of fragmented, hard-to-verify support information by creating a tool that helps users quickly discover real local resources and ask grounded questions about them. In the context of generative AI, we felt it was important to build something that uses AI responsibly: not for vague advice, but for helping people navigate real support options.
Technology Stack:
Languages: TypeScript, SQL, JSON
Frameworks & Libraries: Next.js 15, React 19, Tailwind CSS, Zustand, Zod, LangChain, Gemini, Vitest, Testing Library
Platforms / Cloud Services: Google Maps API, Google Places API, Supabase Auth, Supabase Postgres, Brave Search API
Tools: Node.js, npm, Supabase SQL / dashboard tooling
Product Summary
beacon is a location-aware support navigator that helps users find nearby community services and interact with them through grounded AI. Users enter a location, then receive ranked local results powered by curated data, optional Google Places enrichment, and trusted web discovery. The app includes a dashboard, service detail pages, grounded chat over the loaded service set, a planning flow for longer-term stability, and saved services for signed-in users.
AI Use
Around 90% of the code was generated by AI.
Built With
- brave-search-api
- gemini
- google-maps
- nextjs
- supabase
- typescript


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