gsb. [email protected]
// PHYSICIAN-LED HEALTH IT CONSULTING
gsb. Health IT

Healthcare interoperability and clinical AI, led by a physician.

Medical Informatician · Médico Informático

A physician who speaks engineering.

Twenty-four years bridging clinical practice and health IT. From hospital floors to interface engines, from clinical guidelines to ontologies and FHIR profiles.

Dr. Guillermo A. Sanz Berney is a bilingual physician-informatician based in Santiago, Chile, with dual US/Chilean citizenship and over a decade of senior roles in health IT product leadership and clinical AI evaluation.

His work spans HL7 v2 and FHIR interoperability, semantic mapping against SNOMED CT and LOINC, clinical workflow validation, and the design of rubrics that test how AI systems handle real clinical reasoning. Selected engagements include 24-hospital HIE integration at Northwell Health, oncology ontologies at Varian Medical Systems, and clinical data operations at Signant Health.

He works the seam where medicine and software disagree most loudly, translating clinical reality into structures engineers can build, and engineering decisions into language clinicians will accept.

Where I'm useful.

Four areas where a physician's clinical eye saves engineering teams from expensive misunderstandings.

01

Healthcare interoperability

HL7 v2 to FHIR migration strategy, interface validation, CDA/CCDA design, and clinical workflow analysis. Bridging legacy messaging and modern APIs without breaking what works.

02

Clinical terminology governance

Semantic mapping against SNOMED CT, LOINC, ICD-10/11, and RxNorm. Reference terminologies, code system harmonization, and the careful work of making data mean the same thing across systems.

03

Clinical AI validation

Physician-led review of AI-generated clinical content. Rubric design, hallucination detection, evidence alignment, and the construction of evaluations that test reasoning, not pattern matching.

04

Clinical product strategy

Discovery interviews, BRD/PRD authoring, UAT scenarios, and pilot validation. Translating clinical workflows into product requirements engineers can actually ship.

Recent writing.

A short essay on why HL7 v3 collapsed in the United States and why ICD-11 may follow it, with implications for FHIR-first strategy.

APRIL 2026 · HEALTH INFORMATICS

Two standards, two cautionary tales.

HL7 v3 and ICD-11 in the United States healthcare standards landscape

11 pages · PDF

Download

Let's talk about your data problem.

Independent consulting engagements, advisory roles, and clinical AI evaluation collaborations. Bilingual in English and Spanish. Based in Santiago, working globally.

[email protected]