How Agentic AI and Digital Twins Reshape Clinical Workflows With: Dr. Harvey Castro
Hematologist-oncologist host Dr. Sanjay Juneja sits down with emergency physician and AI advisor Dr. Harvey Castro to demystify agentic AI in healthcare. Castro explains how clinician and patient digital twins could safely scale expertise, reduce ER bottlenecks, and improve decisions when paired with human oversight. The discussion compares build-versus-buy choices for ambient scribing and other use cases, notes when open-source models can lower costs, and outlines how federated learning lets hospitals benefit from broader data without exposing PHI. They also tackle workforce shortages, forecasting near-term wins from ambient documentation, monitoring at home, and culturally aware guidance at the bedside. Throughout, Castro stresses pragmatic guardrails—HIPAA, FDA pathways, and clinician validation—to mitigate hallucinations and bias. For leaders planning an AI strategy, the conversation offers a clear, reader-first roadmap that highlights ai agents in healthcare while staying focused on patient outcomes.
Episode Contents:
About the Guest
Dr. Harvey Castro is an emergency physician, entrepreneur, and Medical AI advisor. He partners with organizations to design safe, practical deployments of AI at the point of care.
Key Takeaways
- Use agentic AI and clinician digital twins to triage, consult, and surface blind spots with human verification
- Decide build vs buy by total cost, customization needs, and data control; open-source can cut ambient scribe costs
- Adopt federated learning to share model gains across sites without sharing PHI; validate locally for safety
Transcript Summary
Q: What is agentic AI and how could it work in care?
A: Agentic AI acts as an autonomous helper. Castro describes digital twins of clinicians and patients that can exchange context, summarize charts, and suggest plans while clinicians verify. He emphasizes warmth and cultural nuance alongside accuracy.
Q: Will agents actually speed ED throughput?
A: Yes, by offloading consults, surfacing latest guidance, and coordinating tasks during procedures. Agents can run overnight, produce transcripts, and flag uncertainties for next‑day review.
Q: Build vs buy for AI in hospitals
A: If resources are limited or needs are generic, buy; if customization and data control matter (for example, ambient scribing), building with open-source models can cost less. Ensure cross‑functional buy‑in and design for the patient as end user.
Q: How to learn from others without sharing PHI
A: Federated learning lets hospitals train on local data and share only model updates. Each site benefits from broader patterns but applies human oversight and local adjustments.
Q: Addressing workforce shortages without replacing people
A: Castro expects human plus AI teams with robotics for lifting and monitoring to extend capacity. He predicts more in‑home care, adherence nudges, and longitudinal context to reduce avoidable visits.
Q: Near‑term wins clinicians can deploy
A: Ambient scribing to fix documentation burden; culturally aware communication aids; consented emotion/micro‑expression cues to time and tailor discussions; after‑visit transcripts for families.
Q: How newcomers should start
A: Be a lifelong learner. Follow diverse AI voices, try tools for your workflow, and evaluate them against patient outcomes. Guardrails like HIPAA and FDA review remain essential.
Q: Risks and mitigations
A: Hallucinations and bias decline as data improve and domain models mature. Keep clinicians in the loop, measure against current standards, and validate locally.
More Topics
- AI in Patient Care
- AI in the Healthcare Industry
- AI and Medical Innovation
- Healthcare Ethics and Policy
Keep Exploring
About the Series
AI and Healthcare—with Mika Newton and Dr. Sanjay Juneja is an engaging interview series featuring world-renowned leaders shaping the intersection of artificial intelligence and medicine.
Dr. Sanjay Juneja, a hematologist and medical oncologist widely recognized as “TheOncDoc,” is a trailblazer in healthcare innovation and a rising authority on the transformative role of AI in medicine.
Mika Newton is an expert in healthcare data management, with a focus on data completeness and universality. Mika is on the editorial board of AI in Precision Oncology and is no stranger to bringing transformative technologies to market and fostering innovation.