Smarter diabetes support, guided by your care team.
A three-year research project co-designing an AI-supported messaging system for adults with type 2 diabetes. The AI drafts personalized, evidence-based nudges on medication, diet, activity, and follow-up care, and every message is reviewed and sent by the patient's own primary care team, keeping the trusted clinician relationship at the center.
- Risk-stratified nudges powered by EHR-based prediction models
- Bilingual (English/Spanish), culturally tailored, and co-designed with patients
- Clinician-reviewed: no message reaches a patient without human approval
From data to doable steps.
A three-part pathway developed with patient advisors and primary care teams at NYU Langone-affiliated practices.
Identify
EHR data flags patients at moderate or high risk of diabetes complications.
Draft
The AI generates brief, guideline-aligned nudges tuned to each patient's language, literacy, and barriers.
Review & deliver
Primary care staff approve and send messages through familiar channels like the patient portal.








