AIPA

Inspiration

Physician visits are often long and tedious, starting much later than the appointment time and sometimes lasting more than an hour. We wanted to create a system that reduces this wasted time for both patients and physician without compromising on information transferred from patient and physician.

What it does

AIPA uses voice-to text and text-to-voice technology to create a webapp that collects information from patients with a friendly, conversational interface.

How we built it

The frontend is built using a React.js library, and the backend is built using Python.

Challenges we ran into

Due to the multiple AI technologies we used in our backend, we faced challenged meshing their functionality with each other and the front end.

Accomplishments that we're proud of

The conversational aspect of the app is very seamless and has near-zero latency.

What we learned

From this project, we gained understanding of web sockets and how to utilize asynchronous calls to external APIs

What's next for AIPA

In future updates, AIPA could implement an algorithm trained on common illnesses and add suggestion to the email send to the physician.

Built With

• deepgram
• groq
• next.js
• python
• react
• typescript
• websockets