ACMS - Testing CAR.L.L., an AI agent for cardiovascular health.

Our lab is seeking motivated and curious undergraduate researchers to work on AI agents and digital twins for cardiovascular simulation. Specifically, we are looking for dedicated and resilient students to contribute to a project funded by the BELS initiative at Notre Dame focused on AI agents for health.

You will be responsible for testing the accuracy of the CAR.L.L. AI agent under various scenarios. These may include quantifying CAR.L.L.'s accuracy in predicting cardiovascular conditions from clinical EHR data, testing its ability to reason based on hemodynamic or fundamental fluid dynamics principles, and evaluating its planning capability for interacting with external simulation tools. You will also be responsible for maintaining detailed records of your interactions with the goal of curating a validation dataset.

Name of research group, project, or lab
The Schiavazzi Lab
Website
https://strategicframework.nd.edu/news/major-investments-announced-to-accelerat…
Why join this research group or lab?

The Schiavazzi Lab conducts methodological and applied research, including:

  • Cardiovascular modeling at various fidelities (0D, 1D, and 3D solvers).
  • Data-driven methods for model synthesis, emulation, and the solution of inverse problems.
  • Uncertainty quantification and multi-fidelity information fusion.
  • Scientific AI agents that combine generative models with physics-based predictions.

If you are curious of what previous undergraduate students have accomplished, a publication is available at this link.

Logistics Information:
Project categories
Aerospace and Mechanical Engineering
Applied and Computational Mathematics and Statistics
Computer Science & Engineering
Information Technology, Analytics, and Operations
Student ranks applicable
Junior
Senior
Graduate Student
Student qualifications

Preferred skills:

  • Basic knowledge of the Python programming language.
  • Familiarity with interacting with large language models.

Students may also opt to audit a course on "Cardiovascular Modeling and Uncertainty Quantification" during the Spring 2026 semester, where many relevant concepts will be explained in detail.

Hours per week
1 credit / 3-6 hours
2 credits / 6-12 hours
Compensation
Research for Credit
Number of openings
5
Techniques learned

The activities will focus on testing the accuracy of the CAR.L.L. AI agent under various scenarios:

  • Predicting cardiovascular conditions from clinical EHR data.
  • Evaluating the correctness of reasoning based on hemodynamic or fundamental fluid dynamics principles.
  • Assessing planning capabilities for interacting with physics-based tools.
  • Curating testing and validation datasets.

Students will gain familiarity with the following:

  • Cardiovascular physiology and health data.
  • Reasoning, planning, and tool use in scientific AI agents.
  • Concepts and tools in cardiovascular simulation.
Project start
Beginning of the Spring Semester 2026
Contact Information:
Mentor
Daniele Schiavazzi
dschiava@nd.edu
Associate Professor/PI
Name of project director or principal investigator
Daniele Schiavazzi
Email address of project director or principal investigator
dschiavazzi@nd.edu
5 sp. | 2 appl.
Hours per week
1 credit / 3-6 hours (+1)
1 credit / 3-6 hours2 credits / 6-12 hours
Project categories
Applied and Computational Mathematics and Statistics (+3)
Aerospace and Mechanical EngineeringApplied and Computational Mathematics and StatisticsComputer Science & EngineeringInformation Technology, Analytics, and Operations