FHIR resources as Delta tables

DeltaForge parses FHIR JSON, XML, NDJSON, and RDF/Turtle resources directly into Delta tables. Query clinical data with SQL. Your data stays on your infrastructure.

JSON, XML, NDJSON, and RDF/Turtle encoding formats
Common FHIR R4 and R5 resources
Runs on your infrastructure; PHI never leaves your network
FHIR Patient (JSON) { "resourceType": "Patient", "id": "pat-12345", "name": [{ "family": "Smith", "given": ["John"] }], "birthDate": "1985-04-12" } FHIR Observation (XML) <Observation> <status value="final"/> <valueQuantity .../> </Observation> parse DeltaForge FHIR parser R4 / R5 store healthcare.fhir_patients id | family_name | given | birth_date | ... Query with SQL SELECT id, family_name, birth_date FROM healthcare.fhir_patients WHERE birth_date > '1980-01-01'

Resource categories

Common FHIR R4 and R5 resources parsed into Delta tables

Patient and encounter

Patient, Encounter, EpisodeOfCare, Practitioner, Organization, Location.

Clinical findings

Observation, Condition, Procedure, DiagnosticReport, AllergyIntolerance, Immunization.

Medications

Medication, MedicationRequest, MedicationAdministration, MedicationDispense.

Care planning

CarePlan, CareTeam, Goal, ServiceRequest, Appointment.

Financial

Claim, ClaimResponse, Coverage, ExplanationOfBenefit.

Documents and bundles

Bundle, DocumentReference, Composition, Consent.

Encoding formats

Parse FHIR resources from any standard serialization

JSON

application/fhir+json, including nested extensions and reference resolution.

XML

application/fhir+xml with full namespace and attribute handling.

NDJSON

Newline-delimited JSON for bulk export from EHR systems.

RDF/Turtle

Linked-data format for semantic web and terminology integration.

Frequently asked questions

FHIR to SQL in practice

How do I query FHIR resources with SQL?

Point DeltaForge at FHIR files in JSON, XML, NDJSON, or RDF/Turtle. Each resource type is parsed into a Delta table with typed columns, so you can join Patient to Observation through the subject reference using ordinary SELECT and JOIN syntax. No Spark cluster and no FHIR server are required for analytics.

Can I land a FHIR bulk export directly?

Yes. Bulk exports arrive as newline-delimited JSON, one resource per line. DeltaForge reads NDJSON natively, so the files an EHR exports become queryable Delta tables without an intermediate conversion step.

Does PHI leave my environment?

No. DeltaForge runs on infrastructure you operate. FHIR parsing, Delta writes, and SQL execution all happen inside your network, which keeps clinical analytics within your existing compliance boundary.

Can FHIR tables be joined with other data?

Yes. Parsed FHIR resources are ordinary Delta tables. Join them to HL7 v2 messages, X12 EDI claims, or any other table in the lakehouse with standard SQL.

Further reading

FHIR Analytics Without Spark walks through landing and querying FHIR data end to end, HL7 to SQL Without an Interface Engine covers the v2 side of the same engine, and Delta Lake GDPR Right to Be Forgotten in SQL shows how to honor erasure requests on patient-level tables.

Bring FHIR into your lakehouse

Parse clinical resources directly into Delta tables on your own infrastructure.

Get Community License Read Documentation