Powering Chart Abstraction Across An Institution: A Celebration of 100 Projects at Vanderbilt

This summer, Brim hit an exciting milestone: 100 research projects completed at Vanderbilt University Medical Center (VUMC) using our AI-powered chart abstraction platform.

From cancer clinical trial pre-screening to analyzing orthopedic injuries from pickleball, these projects represent the incredible breadth of research happening at Vanderbilt, and the power of a tool that’s flexible enough to serve them all.

We’re proud to be deeply embedded in the Vanderbilt research ecosystem, and we wanted to share how this partnership grew, why it works, and what it means for the future of chart abstraction across institutions.

‍

Key Takeaways

‍

• Brim has supported 100 diverse research projects at Vanderbilt, spanning oncology, drug repurposing, sports injuries, emergency surgery and more.
• Our success comes from domain flexibility. Researchers define their own data variables while Brim’s AI and integrations make setup fast and adaptable.
• Vanderbilt has made Brim a core service available to all affiliated researchers, enabling IRB-approved projects to get up and running quickly.
• This collaboration shows how chart abstraction can be democratized at scale, making research 90% faster for anyone in the institution.

‍

Building Momentum to 100 Projects

‍

The Brim team has deep roots at Vanderbilt University Medical Center.

Brim’s founder, Dan Fabbri, is an Associate Professor of Biomedical Informatics at Vanderbilt, where he’s spent years helping researchers use technology to work more efficiently. That connection meant we understood Vanderbilt’s research culture from the inside, and how to build a tool that met their needs.

We were also honored to be selected as an awardee of the Vanderbilt Innovation Catalyst Fund, a program that supports high-potential projects with real-world impact. This recognition gave Brim both credibility and early momentum within the Vanderbilt research community.

Over time, that momentum grew into a diverse portfolio of 100 completed projects, including:

• Cancer clinical trial pre-screening: quickly surfacing potentially eligible patients from unstructured notes.
• Drug repurposing studies: identifying patterns in EHR data that could suggest new therapeutic uses.
• Orthopedic injury research: yes, including tracking the growing number of pickleball-related injuries!
• Emergency surgery management: analyzing cases to inform faster, better care pathways.

The variety of these projects is no accident; it’s a reflection of Brim’s domain flexibility.

‍

Brim is Built to be Domain Flexible and User Friendly

‍

Chart abstraction projects are rarely identical. Every study has its own set of inclusion criteria, relevant data fields, and nuanced definitions. Brim was designed from the start to handle that diversity without sacrificing usability.

Here are a few of the ways we manage that:

Researchers choose their own data.
Brim doesn’t lock you into predefined datasets. Researchers select the records and data sources that matter for their question, whether that’s decades of oncology notes or a focused subset of sports medicine visits.

Researchers define their own variables, with Brim guiding them.
With Brim, you can describe the data points you need in plain English, and our platform will draft the precise AI prompts and structured options to extract them. Over time, researchers can refine and validate those prompts for even higher accuracy.

Seamless integration with existing tools.
For many Vanderbilt teams, that means REDCap. Brim integrates directly with REDCap, allowing researchers to import a REDCap data dictionary, run chart abstraction in Brim, and export results right back to REDCap.

Powerful AI under the hood.
We leverage an institution's generalist foundation model combined with in-context learning to adapt to each project’s unique needs. That’s why the same platform can power a pediatric cardiology study and a neurosurgery outcomes project equally well, and it's one of the many benefits of using foundation models.

‍

A Huge Thank You

‍

We want to take a moment to thank the entire Vanderbilt community - researchers, administrators, and innovation leaders - for making this milestone possible.

By investing in Brim as a core research service, Vanderbilt has shown what’s possible when institutions embrace modern, AI-powered tools: more projects, faster timelines, and data that’s ready for action.

We’re excited to see what the next 100 projects will bring.

If you’re interested in setting up an institutional Brim instance at your organization, whether you’re a large academic medical center or a specialized research network, get in touch. We’d love to explore how we can help make chart abstraction faster, easier, and more impactful for your team.

‍