Jaya Narain

Jaya Narain

Sr. Research Scientist, Apple Health AI

I am a Senior Research Scientist at Apple Health AI, where I work on applied machine learning with multimodal data, with deep expertise in sensor and speech modeling and affective computing.  I was a technical leader for the motion foundation model in Air Pods Pro 3 and also worked on developing and deploying personalized speech models for users with atypical speech for the Adaptive Voice Shortcuts feature.   My research includes characterizing speaking style like naturalness and tone and fairness implications for users with atypical speech, domain-driven methods for learning from motion data, and understanding the mechanics of LLM instruction-following.  

I received my PhD from the MIT Media Lab in the Fluid Interfaces Group, where I studied nonverbal communication in people who don't use typical speech. I was an Apple PhD Scholar in AI for Health & Wellness and an NSF Graduate Research Fellow. I co-founded the MIT Assistive Technology Hackathon (ATHack) in 2013 and led it for eight years, bringing together over 500 engineers, designers, and community co-designers to build assistive technology collaboratively.

My technical foundation spans signal processing, large-scale model training, dataset creation, and model evaluation across both research and product settings. I hold an MS and BS in Mechanical Engineering from MIT, which complements my expertise in human-centered machine learning with a deep understanding of physics and system design.

CV ↓ Google Scholar ↗ LinkedIn ↗

Selected Publications

Full list on Google Scholar ↗

Invited Talks

Patents