Out of the FHIR Podcast · Latest episode Nov 3, 2026
I'm Eugene Vestel. The guides and open-source tools here come from shipping this work: HealthClaw for putting guardrails between agents and clinical data, Open Quality for CMS measures that actually execute. No signup. When the reading runs out, I consult.
Built b.well's first FHIR team · UPMC Health Plan quality analytics · NCQA FHIR advisor, 2025 · 15+ years in healthcare data
Figure 1
What “quality measurement” looks like when it stops being a slide.
valueset "Systolic BP": '2.16.840.1.113883.3.526.3.1032' define "Numerator": exists ( [Observation: "Systolic BP"] BP where BP.effective during "Measurement Period" and BP.value < 140 'mm[Hg]' )SELECT DISTINCT p.patient_idFROM patient AS pJOIN observation AS o ON o.subject_ref = p.patient_idJOIN valueset_member AS v ON v.code = o.code AND v.system = o.code_systemWHERE v.valueset_oid = '2.16.840.1.113883.3.526.3.1032' AND o.effective BETWEEN :period_start AND :period_end AND o.value_quantity < 140Track record
b.well Connected Health
Built the first FHIR team
UPMC Health Plan
5 years · Quality Analytics
NCQA
FHIR advisor · 2025
SQL on FHIR
HL7 IG participant
HL7 FHIR DevDays
Speaker · HealthClaw
HealthClaw
Creator · open source
Standards
I hired and led b.well Connected Health's first FHIR team, stood up Analytics on FHIR there, and mapped the data for the interoperability platform now behind Samsung Health and Fitbit's medical records. The standards work below is where that experience keeps getting tested against other implementers.
Working groups
Speaking
Podcast
Conversations with the people building healthcare interoperability, and the stories behind the standards.
Browse all episodesLatest episodes
Tech Bros Can't Talk to Clinicians: A Payer Interop Engineer's Truth Nuke
Nov 3, 2026
HTI-5 Is a Cleanup. The Real Rule-making Is Happening in Verona.
Oct 27, 2026
The Healthcare Brain: What Vercel’s Agent Numbers Mean for Health Data Infrastructure
Oct 19, 2026
Medicaid Data Got Usable. Nobody Noticed.
Oct 15, 2026
Newsletter
A newsletter on FHIR implementation, quality measurement, healthcare AI, and the tools and standards shaping interoperability. Written for the people doing the actual work.
Written from implementation work. Free to read.
Open Source
A trusted framework for putting AI agents on health data, a personal care agent built on top of it, an open corpus of the quality measures the whole industry keeps rewriting from scratch, and an audit of the federal provider directory. All open source, all running, all usable without hiring anyone.
A security layer between AI agents and clinical data. Redacts PHI on every read, enforces multi-step human approval for clinical writes — proposal, permission evaluation, HMAC confirmation, immutable audit log. Redaction is a Safe-Harbor-style compensating control, not a HIPAA compliance determination; the known gaps are published on its security page.
Spin up a personal health agent in under a minute. Every read redacted, every access audited, every action approved by you — guardrailed by HealthClaw. This is what the guardrail layer looks like once a person is actually using it. In beta, running on synthetic records only.
52 CMS eCQM packages for the 2026 reporting year, with provenance verified in CI and a CLI that validates and packs them. Alongside the measures sits a typed record of what implementers have actually learned about each one. CQL and SQL next to the measure, not buried in a PDF.
An audit of how accurate the federal provider directory actually is, scored across every state and specialty on six dimensions. Click through to verify an individual NPI against the NPPES registry, the OIG exclusion list and SAM.gov, with no login in the way. Scoring is mapped to the REAL Health Providers Act.
More experiments in progress — Curatr Skills and others
See all projects on GitHubStart here
Access patterns, SMART v2 scopes, the redaction boundary people miss, and audit records that survive review.
CQL worked through real CMS logic, plus an open corpus of 52 eCQM packages with CQL and SQL side by side.
Reference architectures and profiling guidance cover the general case. For one decision with real stakes, book a conversation.
Advisory
Fixed-scope reviews with written deliverables, for when the free reference covers the general case and you need it applied to your data, vendors and deadline.
Free 30-minute intro call. Bring one decision you are facing.