Fixing Healthcare With Predictive AI

Guest: Mariano Garcia-Valiño — engineer and healthcare founder (3 exits; now building his fourth)

Episode Summary

Healthcare is burning cash and patience. Mariano lays out a blunt playbook: aggregate real-world signals (labs, pharmacy fills, wearables—even spending patterns that hint at adherence), run AI to flag risk early, and route people to care before conditions explode in cost. No sci-fi. No diagnosis claims. Just practical prediction, consent-driven data, and measurable outcomes.

Key Takeaways

• Cost crisis ≠ destiny: US costs outpace inflation; prevention and earlier intervention are the only scalable fix.

• Data > drama: EHRs + labs + pharmacy + wearables + behavioral/financial breadcrumbs create a far clearer risk picture than any single stream.

• AI’s role today: Triage and risk flags—not final diagnoses. Models surface “high suspicion” and hand off to clinicians.

• Privacy is the moat: Strict consent, separation from employers/insurers, and legal walls keep PHI protected and trust intact.

• What signals matter: From basic blood panels and pharmacy gaps to face-scan metabolic cues—more signals = better precision.

• Why consumers care: Earlier answers, fewer nasty surprises, and lower lifetime spend. Prevention is the new ROI.

• Business model reality: Think subscription + outcomes, not one-off tests. The value is longitudinal.

Chapters & Timestamps

• 00:00 — Why AI in healthcare actually matters now

• 00:34 — Meet Mariano: engineer → serial healthcare founder

• 06:08 — The cost curve problem and why prevention is unavoidable

• 07:44 — What this company really does: navigation + prediction, not diagnosis

• 08:01 — Remote monitoring basics: from wearables to at-home capture

• 09:13 — The messy truth: fragmented data, privacy laws, and integration

• 12:19 — How data flows in (and why employers never see it)

• 13:39 — Why financial/behavioral signals boost predictive power

• 18:13 — What AI tells you: ranges, suspicions, and next clinical steps

• 19:29 — The “why now” for consumers: earlier lifestyle change, lower costs

• 21:10 — Roadmap & what has to be true for this to scale

Notable Quotes

• “We raise a flag, then route you to the right clinician. That’s how AI actually saves money today.”

• “If you stop filling your medication three months in, the model will catch it—and that’s the moment to intervene.”

• “Prediction beats reaction. Every time.”

Practical Uses (Mark’s Playbook)

• Health plans & clinics: embed risk-flag APIs into care navigation and care-gap workflows.

• Employers: fund prevention programs without ever touching employee PHI—measure outcomes only.

• Startups: focus on data rights + consent UX; it’s the difference between demo-ware and deployment.

Call to Action

If you’re building AI for real people—not hype

subscribe to AI Marketing, and DM me on X/LinkedIn @markfidelman If you want the executive playbook on AI agents, join the waitlist at Agentized.com