Inspiration
A 2018 Stanford Medicine research poll found out that on average, PCP (primary care providers) spend 62% of time devoted to outdated, retro EHR software rather than focusing on their primary job: interacting with patients. We believe that we can reform and recreate this outdated system by integrating modern technology such as AI to reduce the time spent on the computer and allow physicians to focus more on the patient.
What it does
BEHR is an improved EHR software that integrates artificial intelligence to optimize the administrative tasks performed by PCP. BEHR has an improved UI that utilizes modern frameworks and design techniques to reduce clutter and promote intuition when navigating EHRs.
How we built it
For our front-end, we utilized the widely popular JavaScript framework React to build our modular UI/UX. In our back-end, we combined the lightweight database Cloud Firestore with the lightweight framework ExpressJS to power our lightweight API. Most importantly, we utilize the modern speech-to-text software AssemblyAI to convert physician-patient conversations into text and Google's cutting-edge LLM model Gemini to create after-visit summaries and tasks for non-physicians to complete in the PCP's place.
Challenges we ran into
We had to learn most of the technology on the spot. We had limited experience in React and no prior experience working with AssemblyAI and Gemini.
Accomplishments that we're proud of
We're proud of completing our project and incorporating all the technologies we set out to do. We're mostly proud of our ability to learn as we build and our perseverance (we spent 20 hours straight working on this project!)
What we learned
As our first hackathon, we learned the enjoyment of programming with our peers in a competitive yet collaborative environment. We also learned our ability to work under pressure as we race for completion before the deadline.
What's next for BEHR
We plan to further align BEHR with HIPAA regulations and improve security. We also plan to further standardize our data to allow communication across different hospitals and health care providers.
Built With
- assemblyai
- css3
- express.js
- firestore
- gemini
- html5
- javascript
- node.js
- npm
- ollama
- react
- render
- server-side
- websockets
Log in or sign up for Devpost to join the conversation.