Yash Ravipati

Clinical AI & Medical Imaging · Harvard Medical School

Most of my research lives here, automating clinical workflows that are tedious and slow by hand. The main project has been sarcopenia assessment from CT scans: figuring out whether a patient is losing muscle mass, which turns out to be a strong signal for how they'll respond to cancer treatment. That system is now being tested in real clinical settings to help guide treatment decisions. I've also worked on foundation models for brain MRI analysis and tumor segmentation in PET/CT.

Brain-Computer Interfaces · UCLA

The focus here is on machine-learning algorithms for BCIs, specifically trying to restore communication for patients with brain stem injuries. The work has also branched into predicting freezing of gait in Parkinson's and detecting epileptic seizures, and I've had the chance to present some of this at international conferences.

Data Science & Automation · Harvard Business School

Prof. Victoria Ivashina

A bit different from my usual work. This was about understanding how private equity deal structures have been shifting, particularly around subscription lines and leverage. Spent a lot of time building out data pipelines and statistical models in STATA to make sense of large PE datasets, which ended up meaningfully speeding up the research workflow.