Orange hoodoos and pines in Bryce Canyon

Kevin Boehm

building AI for cancer biology

Kevin Boehm on a mountain trail

About

I’m a Radiation Oncology resident and Computational Oncology postdoc at Memorial Sloan Kettering, where I lead the Pathology Data Mining team. My research is supported by an NCI K99/R00 Pathway to Independence Award.

I build AI to advance our understanding and management of cancer. My work integrates DNA sequencing with histopathology—pictures of the tumor under the microscope—to profile each tumor’s biology and determine the best way to treat it. I founded Haversine to accelerate the discovery of cancer therapy.

I value building, presence, and accountability: creating more than I consume, being fully present and helping those around me become stronger, and evaluating my actions by their outcomes rather than my intentions.

Training

Work

Mosaic

Establishing digital pathology as a complement to—rather than surrogate for—sequencing in the context of tumor subtyping. Supported by ICI and aiTDIF. Featured at AACR AI and ESMO AI.

Orpheus

Combining clinical text and pathology slides to identify high-risk breast cancer and generate synthetic images conditioned on risk.

Earlier work

Mussel

Scalable preprocessing for digital pathology. Built with Raymond Lim.

Resonance

Annual science conference connecting Yale researchers with New Haven high schoolers.

Interests