Our client needed to virtualize and combine all clinical, operational, and financial data into a personalized data set. They needed a complete data pipeline and analytics platform built on Spark, Scala, RedShift, AWS, and Mongodb.
Problem Statement
The customer needed help building a customized platform to improve the efficiency and transparency of healthcare providers through a vendor-neutral data platform that can scale millions of patients’ records at ease. It included combining all clinical, operational, and financial data to ensure it supports multiple data analytics views, provided search capability, etc.
Our Approach
We analyzed the requirements, and end-user needs, which includes products, marketing, sales, and domain SMEs to design MVP plans and backlogs from various groups in the organization.We adopted the LEARN (Listen, Enrich, Analyze, Reason, and Reinvent) approach to develop the product end to end, including Requirement gathering, Architecting, Designing, development, testing, and support. Complete data pipeline and analytics platform built on Spark, Scala, RedShift, AWS, and MongoDB. We applied DataOps processes to combine an integrated, process-oriented perspective on data with automation and methods from agile software engineering to improve quality, speed, collaboration, and continuous improvement in data analytics.
We coupled the Architecture and Implementation to accommodate frequently changing requirements — Fully customized and argument-driven data processing pipeline with common exception, notification, and error handling.
• Unified Parsing strategy for all EDI, and EMR files with extendibility, easy maintenance, and
configuration-driven.
• Used OSS libraries to reduce the development effort and test-driven deployment strategy to make sure the quality.
• Leveraged stringent security policies, encryptions, compliance, Vault store, Logging, and Monitoring tools.
The below image shows the Architecture of the Digital Native Healthcare Product.
Business Outcomes
The solution worked across all Cloud providers.
• Provided a Platform Approach to scale Terabyte of data processing per tenant
• Achieved quicker product release for newly onboarding providers by 40%
• Reduced the time to market by 45%