Headshot of Harshith Chennupati, smiling

Harshith Chennupati

ML engineer building AI for wearable biosignals & medical devices

Neurava · Johns Hopkins MS ECE

Epilepsy wearableAES 2026 posterIEC 62304

What I work on

Now · Neurava

AI for a clinical epilepsy wearable

Algorithms & Software Engineer at a Purdue spinout building an FDA-track wearable for epilepsy and seizure monitoring.

  • Sleep/wake algorithm for the clinical wearable, presented as an AES 2026 poster
  • Clinical review pipeline around the algorithm, used by analysts for SUDEP-risk and sleep-apnea review
  • Kotlin & Swift native modules (auth, secure credential storage) for the React Native app

Research · Johns Hopkins I-STAR Lab

3D imaging for spine surgery

  • Adapted NVIDIA’s MAISI 3D diffusion model to ultrasound-to-CT, enabling real-time 2D-in-3D visualization for spine surgery
  • Led multimodal cadaver data collection and shipped 3D Slicer modules for segmentation and synthesis

MS ECE, Johns Hopkins: ML for medical applications, medical imaging, signal processing. TA for Deep Learning for Medical Imaging.

Before

  • Vigo, Biomedical Signal Processing Intern: estimated respiratory rate from ECG and shipped accelerometer posture classification
  • Life Signals, Intern: supported ECG collection and testing for FDA-approved wireless biosensors in hospital settings

Toolkit

IMU · EEG · ECG · EMG · PSGCT · MRI · ultrasoundPyTorchMONAISelf-supervised learning3D diffusionC/C++ · CUDASwift · KotlinReact NativeIEC 62304 / V&V

Always happy to talk health AI

Wearables, clinical ML, medical imaging, and frontier AI for healthcare.

Say hello