Relay311

Inspiration

Relay311 helps cities handle reports like potholes, flooding, graffiti, outages, and safety concerns faster. It supports SDG 11: Sustainable Cities and Communities by making reporting easier for residents and clearer for city teams.

What it does

Relay311 is a voice-based 311 reporting tool and dashboard.

Residents call and describe an issue
The call transcript is saved and analyzed
AI identifies the issue, location, category, and urgency
Similar reports are automatically detected and merged
Operators can view and manage incidents from one dashboard

How we built it

Next.js, React, TypeScript, and Tailwind CSS
Supabase for storing calls and incidents
Vapi for voice calls and transcripts
Gemini for classifying reports
Zod for validation
Vitest for testing

Challenges

Calls can be unclear or incomplete
Telling real reports apart from general questions
Detecting duplicates without merging different issues
Extracting useful details from spoken reports
Building a complete product in hackathon time

What we’re proud of

We built a working end-to-end system, not just a mockup:

Phone-based reporting
Transcript processing
AI issue classification
Automatic duplicate merging
Operations dashboard for managing incidents

What we learned

AI works best when it supports people instead of replacing them. We also learned that clear UX, structured data, and duplicate detection are important for making a city reporting system useful.

What’s next

Maps and location clustering
Routing reports to the right department
Alerts for unresolved issues
Citywide analytics
Integration with real municipal systems
Better handling for urgent reports and non-native speakers

Scalability

Relay311 uses a structured JSON format for incident data. Many major cities in Canada and the U.S. use similar formats, so it could be adapted to existing municipal systems without needing to rebuild everything from scratch.