I’m Aarush Kandukoori, a Statistics, Machine Learning, and Computer Science student at Carnegie Mellon. My work spans machine learning research and building real products in healthcare and education.

I co-founded and lead BUMP, a patent-pending closed-loop wearable atropine pump for bradycardia. It combines continuous heart-rate sensing with automated dosing, developed with clinical guidance from faculty at Harvard Medical School. I also co-founded MoCo Innovation, a STEM-education nonprofit that has brought hands-on programming and engineering to more than 5,000 students across underserved schools in Maryland.

My research focuses on machine learning, with recent work in generative and multimodal models alongside applied ML for medicine. Right now I’m an undergraduate researcher in the CMU Database Group, working on query engines and storage — measuring whether splitting a table’s columns across separate Parquet files is free at query time across the 99-query TPC-DS suite under Apache DataFusion. Before that I worked in labs at Harvard Medical School and Brigham & Women’s Hospital, the Stanford Cardiovascular Institute, and the UMD MIND Lab.

Away from the work, I’ve played piano since I was nine, with most of my hours going to Beethoven. I also paint with oils, run, and lift.

I still wish I could go back in time and fix the clam at 1:08!