AmalaJẹun

Inspiration

Finding authentic Amala spots in Lagos is often harder than it should be. Information is scattered across blogs, social posts, or word of mouth, making it unreliable and inconvenient. We wanted to solve this by creating a trusted, community-driven platform where anyone can submit, verify, and discover Amala locations with ease.

What it does

• Submit new Amala spots with details and photos.
  
• Verify locations through community confirmations and reviews.
  
• Discover nearby spots using a map-based interface and AI agentic intake (chat/voice).
  
• It also autonomously scans the public web for likely Amala spots and queues them for user verification.

How we built it

• Frontend: Next.js for a fast, responsive UI.
  
• Backend: Go with Gin for performant APIs.
  
• Database: PostgreSQL on Neon with geospatial support for mapping.
  
• Auth: Google OAuth + JWT for secure user sessions.
  
• Deployment: Railway for hosting and CI/CD.
  
• AI Assistance: Kilocode, Gemini and Chat GPT to accelerate development under tight deadlines.

Challenges we ran into

• Learning Go from scratch while racing against the hackathon clock.
  
• Debugging deployment issues like 502 errors and port mismatches on Railway.
  
• Managing Google OAuth redirect_uri mismatches during setup.
  
• Handling CORS restrictions between the frontend and backend.
• Using a 2dsphere to store geospatial data in neon db for the location property of spots
• Getting the Agentic Intake to work

Accomplishments that we're proud of

• Delivering a working MVP in under 48 hours.
  
• Implementing secure Google OAuth and JWT-based auth flow.
  
• Successfully connecting a Neon Postgres database with the backend.
  
• Building a foundation for autonomous Amala spot discovery and verification.
  

What we learned

• How to rapidly get up to speed with Go and Gin.
  
• Effective collaboration across frontend and backend under time pressure.
  
• Deploying production-ready services with environment configs and CI/CD.
  
• Using AI tools (Kilocode) not just for snippets but for full architectural scaffolding.
  

What's next for AmalaJẹun

• Adding image storage via Cloudinary or a CDN for user-uploaded spot photos.
  
• Implementing AI-powered auto-discovery of Amala spots from the web.
  
• Expanding beyond Lagos to cover other Nigerian cities and dishes.
  
• Polishing the UI/UX and releasing a public beta for community testing.

Built With

• gin
• golang
• neondb
• next.js
• railway
• vercel