Physician AI evaluator
Dr. Adarsh K. Gupta
Physician · Engineer · AI Evaluator · Habit Scientist
Safe, evidence-based AI that actually changes behavior.
Clinical AI evaluation & safety
For teams building health AI
I work first with AI labs and evaluation teams. The work is physician-led testing of health AI: red-teaming, clinical rubrics, reference answers, and a written account of the safety gaps in a defined use case. Consumer health AI apps are the second audience. I review claims, escalation, and the actions a coaching flow may encourage before a product launches.
Since 2023 I've done clinical evaluation work through Mercor, Outlier AI, Snorkel AI Experts, and OpenTrain AI.
Training
Engineering, medicine, and evidence
My training is in engineering and medicine: a BS in chemical engineering, an MS in biomedical engineering, and a doctor of osteopathic medicine. I have been in clinical practice for more than 25 years, and I am a Cochrane Associate Editor.
I was IRB Chair at Rowan University School of Osteopathic Medicine from 2010 to 2024.
- Physician · Engineer · AI Evaluator · Habit Scientist
- BS, chemical engineering · MS, biomedical engineering · DO
- Cochrane Associate Editor
- 25+ years of clinical practice
- Former IRB Chair, Rowan University School of Osteopathic Medicine (2010–2024)
- Author of The Science of Habits and two other books
- Podcast host: 50+ episodes
- HabitRx (live) · CertaintyGrade — Early access
Behavior-change science
The Science of Habits
In The Science of Habits I write about repeated behavior. For AI builders, that expertise informs how I review coaching flows, nudges, engagement loops, and the evidence behind behavior-change claims.
HabitRx is a live AI-supported habit and behavior change platform. CertaintyGrade is Early access, for evidence appraisal and making uncertainty explicit.
Where it started
A short origin
For years I watched motivated people leave a visit with a plan and come back to the same patterns. The question that stayed with me was not whether they had understood the advice. It was what they did the next day.
That question is why I wrote The Science of Habits, and why behavior is part of how I evaluate health AI. A fluent answer is only the beginning. The next question is what someone may do with it.