Students at the Intersection of Law, AI, and Justice Tackle Medical Debt Through Data

Eight students. Four Georgia Tech colleges. One semester-long project with an uncertain outcome. Led by Scheller College of Business Law and Ethics Professor Charlotte Alexander students from across the Institute came together in the Law, Data, and Design Lab to complete a Vertically Integrated Project during the 2025 Spring semester. One team project addressed a growing crisis affecting some of the nation’s most vulnerable: medical debt litigation.

Armed with a desire to do good in the world, and growing expertise in their current studies at the colleges of Business, Computing, Engineering, and Industrial and Systems Engineering, the students discovered how powerful interdisciplinary collaboration and cutting-edge technology can be in creating social change.

The Law, Data, and Design Lab is the brainchild of Alexander, who from a young age felt a call to serve her community. “I went to law school because I saw law as a tool to look beyond myself and contribute to the greater good,” said Alexander. “I see this as part of my purpose. Being at a public university, I take seriously the responsibility to ensure my research is outward facing, that it reaches beyond academia and helps make the world a better place.”

A Real-World Challenge With Real-World Impact

Medical debt is a leading cause of civil litigation against low-income Americans. Often, patients are sued by third-party debt collectors after hospitals sell off unpaid bills. These lawsuits often result in default judgments against defendants who lack legal representation or are still recovering from illness.

The civil legal system in the United States is fragmented. Many case records are recorded differently across thousands of state and county courts, making it difficult to analyze trends or understand the broader landscape of civil litigation. The Legal Services Corporation, the largest funder of civil legal aid for low-income Americans, launched the privately funded Civil Court Data Initiative (CCDI) through its Office of Data Governance and Analysis to build a comprehensive, standardized database of civil legal cases.

Alexander’s Law, Data, and Design Lab partnered with CCDI to develop a method to extract meaningful data from thousands of image-based court filings in Arkansas. The filings were inconsistent, unstructured, and riddled with confusing analysis-resistant markings like stamps and handwritten notes.

“With the rapid development of AI technology and Large Language Models (LLMs), the ability to turn this unstructured data into structured, usable information has the potential to be improved,” said Logan Pratico, a data engineer at LSC. “The students were essential in helping us determine whether the technology has improved significantly enough that such information can be gleaned effectively.”

The end goal was clear, but the results were never certain. To launch the project, Alexander began the semester by grounding her students in the broader context of their work. After an introductory Zoom meeting with LSC, students began researching both the issue of medical debt and the mission and work of LSC. The Zoom meeting and research helped them understand the real-world stakes of their project and how their data analysis efforts could contribute to addressing systemic legal challenges, even if indirectly.

“This was a hard project,” said Alexander. “This was not a canned classroom exercise. I didn’t know if the students were going to be able to figure something out.”

 A “Horse Race” of AI Models

Before writing any code, the students spent a week closely examining the court documents to understand their structure and inconsistencies. “If you're going to do a project focused on data, you have to understand as much as you can about what you're working with,” said Alexander. This hands-on review helped the team conceptualize potential workflows and align on a shared approach.

The students then transitioned into technical work, dividing tasks based on their academic strengths. Computer science students focused on Optical Character Recognition (the process of extracting text from scanned images), while business students validated outputs and assessed their accuracy. This iterative, interdisciplinary process allowed the team to refine their methods and work to develop a robust AI-powered workflow for analyzing complex legal documents.

The team ran what they referred to as a “horse race” of AI tools, testing models like Gemini, GPT, and Grok to determine which could best handle the noisy, image-based PDFs.

“We experimented. We had a lot of work to do in figuring out which dollar amounts in the documents were important and which were not,” recalled Alexander. “It proved to be a more refined process where the students had to really spend time reading the documents and understanding what they included so they could then effectively prompt the models and craft the steps of the workflow to pull out the information that was relevant and useful.”

Unique Perspectives in Interdisciplinary Action

As the project progressed through the semester, it became clear just how powerful cross-disciplinary collaboration can be. The Scheller College of Business students brought project management, communications, and quality control skills, while computing and analytics students handled the technical heavy lifting. Together, they created an effective feedback loop of experimentation and validation.

From the Classroom to the Courtroom

The students presented their findings to LSC and at Georgia Tech’s Undergraduate Research Symposium, where their work was selected for a poster presentation.

“When we initially spoke with the students, we were interested in isolating judgment amounts from the court PDFs” shared Pratico. “They came back to us and asked if there was interest in obtaining not only these data points, but a variety of other pieces of information that – at the time – I wasn't fully aware were available. The students’ ability to be confronted with a challenge and then say, ‘we can actually do you one better’ stands out among all else.” 

The students have sent the resulting code and documentation to LSC. LSC plans to use the tool to analyze trends and potentially publish a national report on medical debt litigation.

“It's no secret technical advancements seem to hit the public sector and nonprofit organizations last,” said Pratico. “The civil legal sphere is no exception.  Students who are interested in applying their knowledge are critical to ensuring new technologies can be applied to solve central challenges confronting civil legal aid. I hope that partnerships like this help to ensure nonprofits and other organizations working in civil legal spheres aren't left behind as these technologies continue to advance at a rapid pace.” 

Looking Ahead

Professor Alexander’s Law, Data, and Design Lab will continue to engage in real work to solve real problems, a process that will always begin in the classroom with the students.

“It was humbling and deeply rewarding to realize our work could directly support people facing significant legal challenges,” said Brezner. “It’s easy to focus on technical metrics like accuracy, false positives, or token optimization for prompting, but presenting our solution to LSC put everything into perspective. Their gratitude for our progress on a problem they'd struggled with for years, and knowing our solution would enhance their workflows, solidified the value of our efforts and reinforced the impact of our work in supporting Americans with their legal challenges.”

As AI continues to reshape industries, Scheller College’s Law, Data, and Design Lab is proving that technology, when guided by purpose and collaboration, can be a powerful force for good.

Learn More: Law, Data, and Design Lab