Inspiration
Non-adherence to diabetic medication has been estimated to occur in as many as 50% of chronic disease patients, resulting in >$100 billion spent each year on avoidable hospitalizations. Studies have shown that having regular, frequent contact with patients by telephone significantly improved patient outcomes.
What it does
The AI nurse checks in with patients independently, checks that the patients adhere to treatment plans and ensures early detection of potential complications.
How we built it
Whisper, OpenAI, ElevenLabs, Python
Challenges we ran into
Narrowing down initial scope. We settled on diabetes and key actions in the treatment journey. However, similar processes apply to other conditions and other tasks at clinics and hospitals.
Accomplishments that we're proud of
Working initial prototype incl. voice agent, function calling, storing data.
What we learned
Voice agents currently still need refinement for great UX to handle interrupts well and minimize latency
What's next for SuperCare
EPIC integration, additional features for additional actions by the agent
Built With
- fastapi
- openai
- python
- vapi
Log in or sign up for Devpost to join the conversation.