Senior Data Engineer — Healthcare / EHR Data Migration (Python + GCP)

🏥 Senior Data Engineer — Healthcare / EHR Data Migration (Python + GCP) We're a GCP-based healthcare technology company. We're hiring a senior data engineer to help medical practices move onto OpenEMR — an open-source, ONC-certified electronic health record system. Each engagement onboards one practice: consolidating their existing records into OpenEMR using standard healthcare data formats, then validating everything before go-live. You'll build this as a repeatable, HIPAA-compliant data pipeline — designed once, then reused for each new practice. This is a hands-on build-and-run role. ────────────────────────────────────────────── 🧰 Tech stack • Language: Python (expert) • Databases: SQL — MySQL / MariaDB (OpenEMR's large ~280-table schema) • Healthcare interoperability: HL7 FHIR R4 (Bulk Data, NDJSON, US Core / USCDI), C-CDA / CCD documents, REST APIs + OAuth2 • Integration engine: Open Integration Engine (OIE) or Mirth Connect • Clinical code systems: ICD-10, CPT, RxNorm, LOINC, CVX • Cloud: Google Cloud Platform — Cloud Run Jobs / Cloud Composer, Cloud Storage, Cloud SQL, Secret Manager, VPC-SC (HIPAA-eligible, under BAA) • Pipeline design: landing / staging zones, idempotent re-runs, batch + delta loads • AI (optional): supervising AI agents that draft data mappings and flag exceptions — you review and approve We are all-in on GCP — real Google Cloud experience is important. ────────────────────────────────────────────── 🧱 What you'll do 1. Work with each practice's records through standard, vendor-supported healthcare data formats (HL7 FHIR, C-CDA, standard reports) 2. Map source fields to the OpenEMR schema (clean / lossy / manual) 3. Build the Python pipeline + integration-engine channels + load paths (REST API, C-CDA import, database import) 4. Validate & reconcile every batch (counts, rules, known-answer checks — as gating checks) 5. Manage a safe go-live (parallel run, final sync, go / no-go, hypercare, rollback) 6. Deploy and run the pipeline on GCP, and package it for reuse across practices ────────────────────────────────────────────── ✅ Must-have • Expert Python (5+ yrs) — pipeline clients, transforms, validators, loaders • Advanced SQL — MySQL / MariaDB, staging design, dependency-ordered loads with FK integrity • ETL / ELT design — idempotent re-runs keyed on source IDs, batch + delta • HL7 FHIR R4 — Bulk Data, NDJSON, US Core / USCDI, SMART / OAuth2 backend auth • C-CDA / clinical documents — parsing & generating CCDs • REST API integration — OAuth2 client-credentials, batching / rate control • Clinical code systems — ICD-10, CPT, RxNorm, LOINC, CVX • Integration engine — OIE or Mirth Connect (or equivalent) • Solid GCP experience — Cloud Run / Composer, Cloud Storage, Cloud SQL, Secret Manager, VPC-SC • Data-quality engineering — count / rule / known-answer validation + reconciliation • HIPAA-aware data handling — encryption, audit logging, minimum-necessary, BAA-bounded work • Git + documentation discipline — versioned mappings (YAML), runbooks, reusable skill docs ➕ Nice-to-have • Prior EHR / clinical-data migration experience (OpenEMR or another ambulatory EHR — eClinicalWorks a plus) • Agentic AI / LLM tooling — supervising AI that drafts mappings / triages exceptions, and knowing where AI must NOT be trusted • Broader interoperability — HL7 v2 (ADT / SIU / ORU / DFT), X12 837 / 835 awareness • PHP literacy (enough to read OpenEMR source) ────────────────────────────────────────────── ⏱️ Engagement Project-based: ~8 weeks of development within a ~13–15-week project. Remote OK, within a BAA-covered environment. Strong potential for repeat engagements as we onboard more practices. ────────────────────────────────────────────── ❓ Screening questions 1. Describe an EHR or clinical-data migration you owned end to end. 2. How have you worked with HL7 FHIR Bulk Data and/or C-CDA documents? 3. What's your GCP experience? Which services have you run in production? 4. How do you design a data pipeline for safe, idempotent re-runs (batch + delta)?