In an emergency call, every second could be the difference between life and death. Unfortunately, across Canada and the world, 911 dispatchers don't have the resources they need to respond to the public need in time. That's not to mention the strain on the dispatchers, of whom 48% screen positive for common mental health disorders. Predictably, this has tragic consequences.

Delta Dispatch augments the work of emergency dispatchers with AI, helping them do their jobs more effectively and efficiently. It processes caller information when no human is available, ranks incidents off of severity, extracts crucial information for the dispatcher and clusters multiple calls from the same incidents together.

Technologies

NextJS
TailwindCSS
FastAPI
LangGraph
- Orchestrates three tiers of AI agents which process call data
Pinecone
- Stores call information in vector form, allowing semantic similarity comparisons to cluster calls
Redis
- Acts as a priority queue for the most important calls
TanStack Query
- Polls the backend to update the dispatcher UI
Twilio

Features

Call ranking
- Analyzes the content of calls and applies insights to a custom algorithm to determine which incidents require the most urgent support
Transcript
- Provides dispatchers with a transcript of every call, helping them get exact information and file better reports
Information extraction
- Gets crucial data, like incident location, from calls and relays it to dispatchers to help them better coordinate responses
Call clustering
- Uses indicators like location, incident type and time to group multiple calls from the same incident together, preventing dispatch centers from expending valuable resources on one problem