Modernizing Health Claims Analytics with Python & Snowflake

Executive Summary

A leading pharmacy benefits provider initiated a modernization program to move beyond legacy SAS infrastructure, seeking scalable cloud-native analytics that would enable deeper visibility into claims and underwriting logic. The project was designed to mitigate SME risk, reduce infrastructure costs, and support broader ambitions in the pharmacy services space.

Factspan enabled the migration of core analytical workloads to Python using Snowpark on Snowflake, rebuilding backend pipelines to improve performance and laying the groundwork for more agile business experimentation.

About the Client

The client is a large-scale pharmacy benefits provider specializing in insurance and medication management services. Their analytics ecosystem supported key business functions such as rebate calculations, trial claim analysis, and effective rate reporting. As their services expanded, leadership sought to enhance their analytical capabilities to better support strategic growth in the pharmacy services domain.

Business Challenge

The client’s analytics backbone was built on decades-old SAS infrastructure, with over 58,000 scripts and a million lines of code. The organization faced urgent risks:

• Rising infrastructure and licensing costs created scaling bottlenecks
• A sole SME maintained most of the business-critical logic and was nearing retirement
• Web-facing SAS modules lacked portability, making direct migration infeasible

There was limited documentation, making code dependency mapping difficult. With growing ambitions in adjacent business areas, the client required a future-facing foundation to drive analytical innovation.

Our Solution

Factspan employed a phased, automation-led migration approach that emphasized performance, traceability, and continuity:

• Discovery & Classification: Using the in-house SAS Code Analyzer, Factspan segmented legacy code into Data Engineering, Data Science, and Visualization tracks and flagged complexity and custom logic.
• Automated Code Conversion: The SAS Code Converter translated 40–60% of the scripts into Python-ready code, with all edge cases documented for manual refinement. YAML-driven configuration ensured transparency.
• Output Validation: The Data Validator accelerator enabled row and attribute-level comparisons, while the Test Case Generator automated validation of key logic.
• Backend Re-Architecture: Instead of replicating frontend components, Factspan rebuilt backend logic and integrations to support modern data workflows and reduce latency within Snowflake.
• Cloud Deployment: All pipelines were operationalized via DBT on AWS Glue with job orchestration and CI/CD support. Post-migration hypercare ensured operational stability.

This transformation allowed the client to retain business logic while removing dependencies on proprietary systems, bringing them closer to a more agile, partner-ready analytics foundation.

Business Impact:

• 60% faster migration enabled by Factspan’s accelerators
• 45% savings on infrastructure and licensing costs
• Backend performance improvements through optimized Python pipelines
• Eliminated key-person risk through full logic documentation and transfer
• Supported client’s evolution towards broader pharmacy services strategy