6 FHIR Integration Engines That Handle Bulk Data Export Right

6 FHIR Integration Engines That Handle Bulk Data Export Right

FHIR Bulk Data Access, anchored by the $export operation, is how US health systems move population-scale clinical data without one-by-one REST calls. The CMS Interoperability rule made bulk data support a baseline expectation for certified EHRs, and the operations that depend on it (population-health analytics, quality-measure submission, Da Vinci PDex payer data exchange) have grown into real production workloads. The integration engines that handle bulk data well are the ones that survive the realities of long-running jobs, NDJSON streaming, and the rate limits that EHR vendors enforce.

Below are six FHIR integration engines that have a real track record on bulk data export in US deployments, with notes on the fit. The complete guide to FHIR-based EHR development in 2026 sets the context, and our FHIR coverage for US healthcare holds the related material.

HAPI FHIR With Bulk Data Module

HAPI FHIR's bulk-data implementation handles $export operations cleanly for moderate-scale workloads. The library generates the NDJSON output, tracks the async job state, and handles the polling pattern the FHIR Bulk Data spec defines. For US EHR teams operating their own FHIR server, HAPI plus the bulk-data module is the canonical path.

Smile Digital Health

Smile Digital Health has invested heavily in production-grade bulk-data export, including support for group-level exports, system-level exports, and the patient-level exports that Cures Act compliance requires. For US health systems running large bulk-data workloads with vendor support, Smile is on the shortlist.

Google Cloud Healthcare API

The Google Cloud Healthcare API supports $export with output to Cloud Storage, which simplifies the storage and downstream consumption story for bulk-data jobs. For US health systems on Google Cloud, the bulk-data integration with BigQuery for analytics is a strong fit. The trade-off is the cloud-vendor lock-in.

AWS HealthLake

AWS HealthLake supports bulk-data export with output to S3 and integration with the rest of the AWS analytics stack. For US health systems on AWS, HealthLake's bulk-data implementation handles the population-scale workloads that population-health teams need without standing up the server-side infrastructure. The fit is for AWS-aligned teams.

Microsoft Azure Health Data Services

Azure Health Data Services supports FHIR Bulk Data with output to Azure Blob Storage and integration with Azure Synapse for analytics. The platform handles the async job lifecycle and the NDJSON streaming that bulk data requires. For US health systems on Azure, this is the native path.

Medplum

Medplum supports bulk-data export through its platform, with integration into the Medplum Bots framework for downstream processing. The fit is for developer-first teams building digital-health products that need to ingest bulk data from upstream EHRs and process it inside the Medplum platform.

What "Right" Means for Bulk Data

A few things separate engines that handle bulk data well from engines that pass the spec test and fail in production. First, async job durability: when a bulk-data job takes hours, the server must survive restarts and retries without losing progress. Second, NDJSON streaming: the output must stream rather than buffer in memory, or the server runs out of heap on the larger jobs. Third, downstream storage integration: the output has to land somewhere the consuming systems can pick it up cleanly, which usually means a cloud object store with appropriate access controls.

The engines listed above handle these well. Many other engines that advertise bulk-data support handle the happy path and stumble when the job runs long, the network blips, or the output size exceeds expectations.

Recommendation by Workload Shape

For US health systems running large population-health workloads, the hyperscaler-hosted options (Google Healthcare API, AWS HealthLake, Azure Health Data Services) usually fit best because the storage and analytics integration is native. For US health systems operating their own infrastructure, HAPI or Smile is the right path. For digital-health products, Medplum's bulk-data plus Bots integration shortens the path from receiving bulk data to processing it. The Top 7 FHIR integration engines for EHR development in 2026 widens the lens to the broader integration-engine decision; bulk-data fit is one criterion among several.

Sources