Home

Michelle M. Li, PhD

Assistant Professor (Incoming Fall 2026)
Department of Biomedical Engineering
College of Engineering
Carnegie Mellon University
Michelle KG

The MEDLI (pronounced as “medley”) Lab at Carnegie Mellon University innovates medical AI algorithms that generate personalized outputs based on the contexts in which they operate, dynamically adapting their reasoning to new cell types and tissues, medical specialties, patient populations, experimental and clinical workflows, and clinical roles. Their design is grounded in biological and medical principles to minimize the risk of contextual error, where predictions appear reasonable but fail to account for critical context-specific information. By developing frontier medical AI and collaborating with experimental and clinical researchers, we build models that seamlessly integrate across experimental and clinical workflows with minimal contextual error, from modulating cellular trajectories to diagnosing rare and complex diseases to inferring treatment response. Ultimately, we aim to innovate context-aware medical AI that treats patients for whom conventional medicine fails.


We are recruiting motivated undergraduate, master’s, and PhD students✨ Please email your CV, a summary of your research experiences, and a description of your research interests. Individuals from underrepresented backgrounds are especially encouraged to reach out. Read more


Upcoming

🌸 August 3, 2026: We will host our annual workshop on AI for Drug Discovery and Development at NYU with Prof. Kyunghyun Cho, sponsored by Genentech.

🌸 August 4, 2026: Michelle will present a poster on Medea, an AI agent for therapeutic reasoning across biological contexts, at Microsoft Research’s AI Scientist Summer Workshop.

🌸 August 9-13, 2026: Michelle will present GraCE-VAE, a causal representation learning approach for single-cell perturbation prediction, at KDD AI for Science.

More

🌸 September 1, 2026: The MEDLI Lab officially opens!

Recent News

🌸 July 3, 2026: COMPASS, a generalizable AI model for predicting immunotherapy outcomes across cancers and treatments, is published in Nature Medicine and featured on HMS News.

🌸 June 2, 2026: Michelle presented a keynote at NetBioMed on Multiscale Network Medicine: From Epigenomics to Therapeutics.

🌸 May 27-29, 2026: Michelle gave a keynote at the 3rd SciFM Conference on Foundation Models and AI Agents for Science.

⌛️ More

🌸 May 26, 2026: Michelle presented at Imperial College London’s Better Bioinformatics Seminar Series.

🌸 May 9, 2026: Michelle led a tutorial in the “Frontiers in Computational Methods” session at Northeastern University’s May Institute.

🌸 May 6-8, 2026: Michelle organized another excited iteration of SAIL. Check out our NEJM AI perspective on the integration of AI into clinical medicine from last year’s conference. Stay tuned for a follow-up.

🌸 April 24, 2026: Michelle presented CLEF, a controllable sequence editing approach for biological and clinical trajectories, at ICLR.