I bridge the gap between raw healthcare data and bedside clinical reality. I build and audit AI systems that are safe, interpretable, and operationally defensible.
- Clinical ML: Risk stratification, threshold optimization, and precision/recall tradeoffs framed around clinical consequence
- AI Governance: Failure mode analysis, problem scoping, and clinical safety guardrails
- Healthcare Ops: Workflow integration, CMS compliance workflows, and clinical analytics
- Healthcare AI Failure Mode Case Studies: Abstracted audits of where healthcare AI fails at the intersection of clinical authority, workflow, and accountability
- Post-Discharge Follow-Up Risk Model: Python-based predictive model with threshold analysis framed around care team capacity and clinical risk tolerance
I audit clinical AI systems for the failure modes that technical testing cannot catch — behavioral drift, displaced safety nets, and accountability gaps that surface after go-live.
If you're building or deploying AI in a clinical setting and want a bedside perspective on where it may fail in practice, book a discovery call.
Book a Clinical AI Discovery Call →
"AI can predict a readmission, but it can't tell you the patient's daughter is the primary caregiver. I build for the gap in between."