HopHacks 2025

Johns Hopkins University

← Back to Home

Competition Overview

  • Placement: 2nd of 102 teams Overall & 2nd Place in Healthcare track.

Project Links

View the Devpost submission

View the GitHub repository

Technologies Used

  • Redbull (zero sugar)
  • Gemini
  • Flask
  • Python
  • JavaScript

Details

For this project I worked with two new teammates: Jack Deye (UCLA) and Saketh Poori (SFU). We hadn't met before HopHacks, and it was a really fun and collaborative 36 hours. I'm pretty proud of the work we did and we've done our best to keep in touch since, Saketh and I were able to meet up at HackHarvard a couple months later which was a lot of fun.

Team + some engineers at Commure who judged our project

MedRelay was built to reduce the administrative burden that healthcare providers face when completing insurer claim and prior authorization forms. The system listens to natural doctor-patient conversations, transcribes them, and turns the conversation into structured information that can be mapped into standardized medical paperwork.

Agent outputs structured data that is mapped to form fields

A big part of our project was focusing on a human-in-the-loop workflow rather than replacing providers. It extracts key medical details, identifies outcomes like prescriptions, referrals, and tests, and then fills the appropriate form fields so a provider can review and submit a draft quickly with fewer errors.

Redbull :)