Modernize Analytics

A leading healthcare cost containment company focused on optimizing claims and pricing operations sought to modernize analytics capabilities. It needed better access to actionable insights to support strategic decision-making and business growth.

Challenges

The client was facing growing challenges in getting the business insights they needed from their data. In spite of having many Power BI reports and an Azure SQL data mart, which they believed to have the “right data”, they still were not able to get the information that the business needed.  Furthermore, it wasn’t fully clear why their analytics environment was unable to meet business needs.

Our solution

The client first engaged Celsior to do a two-day data strategy workshop to assess data architecture, management, and governance. This built confidence between us and provided a starting point for further understanding the issues with their analytics environment.

Following the data strategy workshop the company engaged us to go through a two-to-four week discovery process to clarify the issues in more detail and put together a plan for resolving them.

We identified two key areas where the analytics ecosystem needed modernization:

• Embedded logic in Power BI: The organization had embedded complex business logic and data transformations directly within
  Power BI reports, leading to redundant and inconsistent logic across dashboards, limited scalability and maintainability, and a lack of centralized data governance.
• Unreliable data infrastructure: The company’s data infrastructure was hindered by unreliable ETL pipelines, frequent ingestion errors, and manual post-processing. The infrastructure also lacked flexibility to handle fluctuating data volumes. These inefficiencies delayed reporting, increased operational overhead, and compromised data quality. The absence of automated validation made it difficult to deliver trusted, client-facing analytics and dashboards—impacting strategic decision-making and stakeholder confidence.

We co-created a process with the client and took the following steps to rationalize and modernize the data and analytics infrastructure.

Centralized analytics data model

• Replaced fragmented Power BI logic with a centralized analytics data model to improve consistency and maintainability.

Pipeline error tracking, data validation, and data quality

• Modernized Legacy ETL – The legacy ETL processes (including NPPES) were rewritten using Python for full and incremental loads, improving data accuracy and refresh cycles.
• Real-time error tracking – ADF was integrated with Azure Monitor and SQL-based logging to capture and instantly alert on pipeline failures.
• Automated data validation – Stored procedures were deployed to verify row counts and flag mismatches, reducing manual checks from hours to minutes.
• Embedded Data Quality Framework – Automated checks were implemented for missing values, duplicates, and formatting issues.  Quality scores and audit-ready reports were generated.
• Azure SQL – Enhancements were made to the Azure SQL data mart, the target of the ETL processes.

Scalability through dynamic pipeline architecture

• Enabled ADF pipelines to auto-adjust for varying data volumes and sources, ensuring consistent performance and scalability.

New Power BI dashboards

• Developed over 40 Power BI reports/dashboards with a reusable, model-driven architecture resulting in deeper insights into claims, sales, provider pricing, and performance metrics.

Automated reporting

• Implemented PDF auto-generation and SharePoint integration for client-facing reports, reducing manual effort and improving delivery speed.

Seamless collaboration & governance

• Maintained bi-weekly reporting cadence and transparent documentation to ensure knowledge transfer and alignment with the client’s stakeholders.

Business outcomes

From an operational efficiency standpoint, the streamlined data pipelines reduced data loading delays by 40 percent and manual exception handling by 60 percent.

User adoption also significantly improved with usage of dashboards and reports by internal stakeholders increasing from 50 to 85 percent.

Most importantly, modernization of the analytics environment empowered marketing with real-time insights, boosting campaign precision and sales performance and contributing significantly to high company revenue growth.