Rescue Remix

Inspiration

Emergency response teams deal with a huge amount of call noise: accidental dials, non-emergency requests, and unclear first reports. We wanted to build something that helps dispatchers focus on critical situations faster, without removing humans from the loop. That idea became Rescue Remix: an AI-assisted triage layer for emergency call workflows.

What it does

Rescue Remix provides a real-time operations workspace for emergency call handling:

AI-assisted voice intake for live conversations
Real-time transcript streaming during calls
Dispatcher dashboard for incoming calls and recent history
Live monitor view for supervisors
Analytics dashboard for outcomes, volume, and trends

How we built it

We built Rescue Remix as an end-to-end web platform with a modern frontend, voice AI integration, and realtime data:

Frontend workspace in Next.js + React for dispatcher, monitor, and admin views
Voice AI integration using ElevenLabs Conversational AI for live call handling
Realtime + persistence with Supabase (auth, database, and realtime subscriptions)
API routes for secure server-side operations (like signed URL handling and conversation fetching)
Analytics visualizations for call outcomes and operational insight

Challenges we ran into

Designing a flow that stays fast under real-time updates
Making live transcripts usable inside dispatcher workflows (not just raw logs)
Keeping the architecture clean across multiple roles (dispatcher, supervisor, admin)
Balancing product clarity with technical depth for a hackathon timeline
Handling imperfect/variable call data while still generating meaningful analytics

Accomplishments that we're proud of

Built a complete end-to-end prototype in a weekend
Integrated voice AI with real-time transcripts into a usable dashboard
Designed a dispatcher workflow that keeps humans in the loop
Turned noisy call data into clear, actionable insights
Delivered a clean, working demo across frontend, backend, and realtime systems

What we learned

Real-time systems require strong UX decisions as much as backend speed
“Human-in-the-loop” design is essential for trust in emergency tooling
A working end-to-end prototype tells a stronger story than isolated features
Integrating voice AI into operational workflows is powerful when grounded in clear user roles
Two members of our team were participating in their first hackathon, which pushed us to prioritize building a complete, end-to-end system