Episode Description
Travis Garland spent two decades in healthcare operations running data analytics for one of the largest health payers in the world before going all-in on AI agents. As CEO of Kickstand AI, he builds agents for medical practices. But during a routine scheduling project, his team discovered something alarming: front-office staff were pulling patient schedules, containing protected health information, and dropping them straight into a public AI prompt to save time.
That moment became the origin story for My Compliance Citadel, the governance layer Travis co-founded with Thomas McShane to keep AI agents in check in healthcare and beyond. In this episode, Chris Hutchins and Travis get into how agents drift from their training over time, why you can't simply prompt your way into better-behaved AI, and the "trusted agent score" (TAC score) that functions like a FICO score for AI performance. They also cover the tangled state of AI regulation in the U.S., 151 different compliance standards and counting and who actually carries the liability when an agent gets it wrong.
In this episode:
- The real story behind the patient data leak that led to My Compliance Citadel
- Why AI agents "want to make you happy", and why that's a problem
- The TAC score: scoring AI agents like a credit score
- Why EHRs shouldn't be the customization layer for AI
- Navigating HIPAA, the EU AI Act, and dozens of overlapping compliance standards
- Why trust, not technology, is the real challenge of the AI era
Chapters:
00:00 – Cold Open: A Rare Vantage Point
01:39 – From Psychology to Healthcare Operations to AI
04:53 – What Real Practices Feel When AI Enters the Room
11:56 – The Ambient Scribe Problem: Three Documents, One Public Prompt
16:59 – Why AI Gives You Choice, Not Just Time
19:06 – The Payer vs. Small Provider AI Arms Race
22:32 – Where AI Actually Earns Its Keep
26:59 – Should AI Live Inside the EHR — Or Above It?
29:13 – The Leak: How a Scheduling Project Exposed Patient Data
30:52 – Building My Compliance Citadel: The Hallway Monitor Model
35:26 – The TAC Score: A FICO Score for AI Agents
41:03 – Why You Can't Prompt Your Way Into a Better-Behaved Agent
46:42 – Navigating HIPAA, the EU AI Act, and 151 Compliance Standards
53:23 – Who's Really Liable When AI Goes Wrong
58:47 – The Bigger Problem: Eroding Trust in Leadership
1:02:43 – The 60/40 Knowledge Gap Facing Junior Staff
1:09:32 – Where to Find Travis and Kickstand AI
About The Signal Room: The Signal Room is a podcast and communications platform exploring leadership, ethics, and innovation in healthcare and artificial intelligence. Hosted by Christopher Hutchins, Founder and CEO of Hutchins Data Strategy Consultants. Leadership, ethics, and innovation, amplified.
Website: https://www.hutchinsdatastrategy.com
LinkedIn: https://www.linkedin.com/in/chutchins-healthcare/
YouTube: https://www.youtube.com/@ChrisHutchinsAi
Book Chris to speak: https://www.chrisjhutchins.com