LifeLine

The page
The thinking

Inspiration

LifeLine was born out of the need to support emergency dispatchers during critical moments. We recognized that every second counts in a crisis, and we wanted to harness conversational AI to help ensure that callers are helped.

What it does

LifeLine is a real-time communication assistant designed for emergency dispatch scenarios.

How we built it

Twilio Media Streams + Assembly AI + Gemini Conversational AI + Flask & Socket IO + Asynchronous processing & Threading

Challenges we ran into

Real-time integration, connecting and passing data around various services, latency issues, reliability in answers

Accomplishments that we're proud of

Achieved bidirectional communication between user and phone, proprietary encoding, and audio chunking algorithm to pass audio seamlessly across APIs

What we learned

We learned about the power of conversational AI, the best practices for real-time data streaming with WebSockets, and how to maintain low latency in applications that bring together various APIs.