From TreeHacks to The Garage: Northwestern CS Team Launches Healthcare Automation Platform

Administrative complexity costs the healthcare industry more than $265 billion annually, according to a 2019 review published by JAMA Network.

This waste in spending is exacerbated in the country’s network of rural hospitals. A 2023 report from the Center for Healthcare Quality and Payment Reform found that more than 30 percent of these hospitals are at risk of closure due to financial instability.

Addressing these challenges, a Northwestern Computer Science student team including Justin Dong, Jiayan Luo, and Isaac Miller, and collaborator Akshay Murthy (Purdue University, Workday) aim to provide a comprehensive suite of services to automate hospital administration tasks such as prior authorization forms and speech transcription.

In less than 48 hours during Stanford University's TreeHacks Hackathon in February, the team built MediFlow, a healthcare platform for rural hospitals. They won first place for “Best Use of Convex Features.” Convex is an open-source backend for application builders and was a sponsor of the TreeHacks event.

Based on the enthusiastic interest the team received from healthcare professionals and accelerators following the hackathon, the group decided to turn the MediFlow project into a startup. The team joined the spring 2024 cohort of resident teams at The Garage at Northwestern University.

Luo, a fourth-year student pursuing a bachelor’s degree in computer science and a minor in history at the Weinberg College of Arts and Sciences, explained that in recent weeks the team has conducted outreach to local hospitals to learn more about administrative complexities in healthcare. They are also exploring entrepreneurship fellowships.

“I had never programmed before college, so going from that to building a functional app was a nice benchmark to assess how much I’ve learned these last few years,” Luo said. “It was really rewarding to combine skills I’ve learned in different CS classes at Northwestern — such as building machine learning pipelines in COMP_SCI 449: Deep Learning, creating full-stack applications in COMP_SCI 396: Full Stack Software Engineering, scaling backends in COMP_SCI 310: Scalable Software Architectures, and general coding practices learned from the core CS courses.”

The team had a fun and meaningful experience at TreeHacks and is excited to continue working on MediFlow.

“It was fun to work with the team to rapidly prototype a large language model application,” said Miller, a fourth-year student earning a combined BS/MS in computer science and a minor in engineering management at the McCormick School of Engineering. “We had to figure out how to evolve what we were making over the course of the hackathon and learn to create adaptable components for the direction we chose.”

“It was exciting and fulfilling to work with Akshay, Isaac, and Jiayan. Their energy and enthusiasm were contagious,” said Dong, a third-year student earning a combined BS/MS degree in computer science at Northwestern Engineering. “Our team loves building new things. Hackathons like TreeHacks cultivate that passion by acting as an accelerator for designing, iterating, and developing concepts.”