S. A. I. Shouborno

My research develops interpretable machine learning methods that infer what sensors cannot measure directly, such as blood pressure from a skin-worn patch or the program execution trace of an industrial controller from the sensors on the plant it controls.

PhD candidate in Electrical and Computer Engineering, University of Massachusetts Amherst. Advised by Bashima Islam.

Open to postdoctoral and industry research roles. PhD expected May 2028.

Portrait of S. A. I. Shouborno
  1. Wearable health sensingHealth
  2. Industrial cyber-physical systemsMachines
  3. Explaining sensor data in languageLanguage
  4. Efficient inference on the deviceOn-device

About

At UMass Amherst I estimate blood pressure from graphene electronic tattoos with the Kireev Lab, and I reconstruct industrial controller behavior from process sensors with Bosch Research. Before that, at WPI, I built LLaSA, a language model that reasons about human activity from IMU data, and JOLT, a training method that keeps early-exit networks within a compute budget. Before my PhD I spent two and a half years shipping speech and conversational AI at Hishab.

Selected papers

All publications

News