epocrates president David Minkin on turning 25 years of clinician trust into an AI-powered clinical intelligence platform without disrupting the workflow.

Episode Resources:

• Overview of epocrates AI Assist and Clinical Decision Support
• Official epocrates Product Page and athenahealth Integration
• Analysis of epocrates’ Role in Reducing Clinician Cognitive Load
• Video Discussion on Ethical Guardrails for AI at the Point of Care

For a generation of physicians, epocrates was the first medical app they ever downloaded. It lived on Palm Pilots before smartphones existed, replaced the dog-eared Physician’s Desk Reference, and became the reflexive check for a dose or an interaction between patients. Twenty-five years and a million daily users later, it is still there. In this episode, Trevor Royce and Tim Showalter dig into what happens when one of the most deeply embedded tools in medicine decides it wants to be something bigger than a lookup app.

They are joined by David Minkin, president and general manager of epocrates, who came to the role from digital media rather than clinical practice. Minkin describes the work as a reinvention problem rather than a repair job, and he is blunt about the constraint that shapes every decision: trust is not an aspiration, it is a hard limit. That principle explains why AI Assist carries no sponsor influence, why the product refuses to diagnose or prescribe, and why 75 clinicians in an alpha lab generated more than 5,000 pieces of feedback before the feature went wide. The conversation covers the shift from information to intelligence, why pull-based engagement beats another alert firing in a busy clinic, and the metrics Minkin actually watches, including how many seconds it takes a clinician to get an answer.

Trevor and Tim also unpack the strategic picture around the company, from the athenahealth acquisition and where a reference tool sits next to an EHR, to what a 25-year curated content library is worth in a market full of AI-first entrants scraping the open web. Minkin closes with pointed advice for health tech founders, arguing that the real opportunity is not autonomous diagnosis but the far less glamorous work of cutting cognitive load for clinicians who are already exhausted from chasing down information that should be instant.