Data Platform API
The most critical feature of the Bulk Data API is self-service, easy-to-use, bulk data access via a “data warehouse” paradigm.
Conceptually, health data flows from data sources into an integrated transactional OLTP data model- the engine of HIEBus. Subsequently, the transactional data is enriched and then subsetted into OLAP data marts hosted in Redshift.
Amazon Redshift serves as the persistence layer for all CE data in an MPP OLAP architecture and makes it available via a JDBC Endpoint.
The Bulk Data API allows applications to “tap into” the pipeline at different levels to provide “transactional data” vs. “analytics-ready enrichments”
Tableau and most other analytics tools support native, high-performance, and high-reliability access to Amazon Redshift data warehouse. Armed with a Redshift JDBC endpoint, credentials, and a data model, an external credentialed user can hook up nearly any analytics tool or development framework without having to figure out how to process files and differential data operations.
The original transaction model and the enriched data model are made available to external and internal data consumers via a high-performance managed data warehouse (AWS Redshift) that is optimized for OLAP style bulk data operations. This data access modality facilitates “self-service” data access to a Care Evolution data platform.
The enriched data model can facilitate several use cases including:
- Direct data access to appropriate subsets to customer-built applications or services.
- Direct data access for third party data analytics tools such as Tableau, Microsoft Power BI, etc. (almost all analytics tools provide native access to AWS Redshift).
- Access via Galileo to support workflows for which the Galileo UI is uniquely suited (i.e. drill down to patient list, subsets for practices).
- Access via additional AWS Services for subsequent processing.
- Fast bulk access to HIEBus services such as Beacon, Bulk FHIR or Standard Data Out.
From a customer/external data consumer perspective: all CE data are all managed in a single, managed data warehouse that is optimized for fast bulk data access., including:
- HIEBus source data concepts such as claims, encounters, procedures, problems, observations, etc.
- HIEBus enriched data concepts such as recordgroups, term mappings and subsets, provider mappings, etc.
- Galileo enriched data concepts such as derivedencounter, participant (in an RKStudio survey), hccprofile, medicationclass, measureresult, riskfactor, etc.
