Allegheny Health Network Pilots AI Platform to Manage Diabetes Between Doctor Visits

Allegheny Health Network will pilot UpDoc's AI platform to monitor diabetes patients between visits, aiming to speed insulin adjustments under physician supervision.
Allegheny Health Network (AHN) is preparing to pilot an artificial intelligence platform designed to help primary care physicians manage patients with Type 2 diabetes between office visits, marking a step beyond the administrative uses that have dominated healthcare AI deployments.
The 16-hospital, Highmark-affiliated health system has partnered with Palo Alto-based UpDoc, according to Pittsburgh Post-Gazette, to deploy a conversational AI platform that communicates with patients through phone calls or text messages, monitors glucose readings and symptoms, and recommends insulin adjustments within treatment parameters established by physicians. Patients whose glucose levels remain outside those physician-defined ranges are directed to contact their care team for further evaluation.
Unlike ambient documentation tools that automate clinical notes, the platform is intended to support ongoing disease management under physician oversight. The approach reflects an expansion of healthcare AI into clinical workflows, following efforts such as Mayo Clinic and Microsoft's work to develop a healthcare-specific frontier AI model.
According to UpDoc, every interaction between the AI agent and the patient is recorded in the electronic health record, allowing physicians to review recommendations and patient responses before making treatment decisions. The company describes the platform as a physician-directed clinical decision support system rather than an autonomous care provider.
"It’s adding a cook to the care team, it’s opening the door and allowing the AI agent to join the team," UpDoc CEO and primary care physician Dr. Sharif Vakili said to Pittsburgh Post-Gazette. "It’s adding something important between doctor’s visits; it’s doing things in a new way."
UpDoc introduced the platform after receiving FDA clearance for its Software as a Medical Device (SaMD) designed for Type 2 diabetes management in June. The company says the software enables physicians to prescribe AI-supported care plans while maintaining clinical oversight.
Clinical Evidence Drives Expansion
The pilot builds on clinical evidence published in JAMA Network Open in 2023, where researchers evaluated an AI-assisted insulin titration platform involving 32 adults with Type 2 diabetes.
The randomized clinical trial found patients using the AI platform reached effective insulin dosing in a median of 15 days compared with more than 56 days under standard care. Medication adherence also improved to 83% versus 50% among patients receiving conventional management. The study attributed delays in insulin optimization to physician time constraints and what researchers described as therapeutic inertia.
Stanford Medicine, where the research originated, said the conversational AI platform was designed to help patients safely adjust insulin between appointments while following physician-approved treatment protocols.
The AHN deployment comes as health systems increasingly evaluate AI for chronic disease management beyond documentation and administrative automation. Similar initiatives are expanding AI-supported remote care programs that monitor patients outside traditional clinical settings.
In a statement announcing the partnership, Amy Crawford-Faucher, Chair of the AHN Primary Care Institute, said the health system expects the platform to strengthen care coordination, reduce administrative burden, and allow physicians to spend more time on clinical decision-making as integration with its electronic health record progresses.
Healthcare leaders continue to caution that patient-facing AI requires careful oversight. Derek Angus, Chief Innovation Officer at UPMC, told the Pittsburgh Post-Gazette that AI could significantly improve home-based disease management but warned that these systems must be validated because incorrect recommendations remain a potential risk.
UpDoc, founded in 2023, has raised $16 million in seed funding from investors and strategic partners including the American Diabetes Association Innovation Fund, Eli Lilly, Mayo Clinic, and Cathay Innovation. The company says it plans to expand its AI platform to additional chronic diseases in the coming months.
Key Takeaways
- Allegheny Health Network pilots AI to enhance diabetes management between doctor visits.
- UpDoc's platform monitors glucose levels and recommends insulin adjustments under physician supervision.
- AI interactions are recorded in patient health records for physician review and decision-making.
- The initiative expands healthcare AI's role from administration to active clinical support.