8 Corporate Park, Irvine, CA 92606

FHIR-Based and Streaming-Enabled Healthcare Big Data Repository

Next Generation Healthcare Data Repository

  • Streaming-Enabled, Big Data-Enabled, Native FHIR schema, and FHIR Interfaces 

Native FHIR Schema

We turn hundreds of FHIR resources into typed relational tables using FHIR RDBMS schema on a big data platform. It enables healthcare applications developers to access data through standard JDBC drivers or standard FHIR RESTful API without worrying about the complexity of big data and without learning about FHIR data modelsWe remove the technology hurdles and complexity so healthcare applications developers can focus what they do best, developing intelligent applications that require big data, more quickly, more efficiently and at a lower cost.   

Streaming Data 

Incoming request messages or response messages are turned into data streams that are stored into typed data tables and raw data lakes, and can be subscribed by a large number of apps. Subscribing real-time analytics apps extracts information or learns from the streaming messages to provide clinical insights.   

Distributed Parallel Computing 

All operations are transparently and automatically divided into small tasks and distributed to multiple nodes of a cluster and executed natively in parallel.  Distributed and parallel computing provides high-performance large-scale throughput at a lower cost, compared with traditional RDMBS systems.   

Massive Scale-up 

Hadoop-based distributed systems are cost-effective for scaling out dynamically for any a large volume of applications.    

Flexible | Lightweight 

A traditional healthcare data repository usually includes hundreds of healthcare entities. Deploying a healthcare data repository traditionally means deploying it across the entire set of hundreds of healthcare entities. Each healthcare entity has no choice but to pay for features they do not need.  Our solution allows customers to deploy healthcare resources incrementally.  We provide flexible options for customers to choose any number of FHIR resources to deploy from hundreds of FHIR resources based on business needs. Customers only pay for what they need.   


Our FHIR RDBMS schema are extensible dynamically, each table and each field of each table are extensible by redefining the default meaning or adding new fields during run time. No schema changes are required after deployment.   

Schema on Write and Read

Our platform is designed to enable both schemas on write and on read for machine learning and data science applications processing different non-structured, semi-structured and structured data. 

Hybrid Lambda Architecture 

Our platform is designed to enable both batch and stream-processing natively for handling massive quantities of healthcare data. It can run both OLTP and OLAP applications. 

FHIR-Based Patient Registry  

A patient registry is a collection of standardized demographical information about a group of patients and services as a single source of truth for patients. The patient registry is a critical module in any healthcare information systems. A traditional patient registry is built on proprietary patient data models and is integrated with a master patient index through a proprietary interface. It is difficult and costly for customers to integrate with healthcare systems, and switch MPI vendors.  With deep and extensive experience in building patient registry products and solutions for different vendors, we provide an innovative next-generation patient registry: 

  • FHIR-Based Patient Model: Standardize the patient data model 
  • Patient Data Model Extension: Allow patient schema extension in runtime by changing the default meaning of fields or adding new fields without requiring re-deploying the schema and update the software 
  • FHIR MPI API: We define a complete set of FHIR MPI API for matching, probabilistic search, merging, unmerging, linking and unlinking. This standard FHIR MPI API simplifies the switching to a different MPI vendor
  • A single source of truth: Generate a single source of truth through configurable strategies
  • Data cleaning and normalizing: Provide built-in data cleaning and normalizing functionality so customers do not need to integrate third-party software for data cleaning and normalizing 
  • Integrate with Leading MPI vendors: Integrate with Oracle healthcare master person Index product by default 
  • Cloud-based or on-premise 
  • Big data platform

FHIR MPI Interface 

  • MPI interface has always been proprietary and vendor-specific. The interface technology used by different vendors is also distinct. Some vendors use SOAP, some vendors use Java, and some vendors use REST. As a result, it is costly and difficult to integrate an MPI solution with healthcare information systems 
  • We define a complete set of FHIR Patient MPI commands as a standard MPI interface. Following the FHIR specification, we use RESTful API as the standard MPI interface. Request and response are defined for each FHIR MPI command.