Miriam Tse-Holloway

Small practices face the same risk analysis mandate as large health systems, with no exemptions.

Deploying AI in medical practice triggers existing HIPAA training obligations staff may not know.

AI tools require a BAA based on how they handle patient data, not their marketing claims.

These certifications answer different questions and leave different gaps in healthcare AI.

Errors go live in systems, not draft queues, changing what mistakes cost a practice.

Healthcare systems need workflow-specific metrics to evaluate whether AI agents actually work.

RPA tools fail silently in EHR systems because they depend on unstable interfaces.

Three concrete fears drive staff resistance, and administrators can address each one.

Healthcare AI agents need audit logs built to capture decisions, not just access.

Three channels reach different patients at different times to stop no-shows before they happen.

Missing deadlines kills claims with no appeal path available.