For over fifty years, this firm has provided auditing, business planning, financial services, and expertise on employee benefits. To better serve their customers, this firm needed to transform their existing analytics and reporting capabilities to allow analysts to study larger, external datasets from multiple sources to project the best outcomes. The original process, executed manually, was inefficient and caused bottlenecking. Using a Data Platform in 30 Days (DPi30) engagement, funded by Microsoft, Tallan built a modern data platform pilot as a solution to pull CMS (Medicare) data from disparate applications and sources. The team automated data processing using Azure tools, improving efficiency, and reducing errors.
Cloud, Data & Analytics
Define, Ingest, Automate, and Report.
- Tallan interviewed the client’s team to realize use-cases, identified ingesting substantial Medicare data files as the priority for the pilot, and worked with the client to understand the technical requirements for implementing the new Azure data architecture.
- Tallan began by analyzing the current data estate. From these findings, Tallan defined the manual, unstructured process for data ingestion and reporting and confirmed that a basic modern data platform design underpinned with Azure Data Services would be achievable within the one-month engagement timeframe.
Decreased Processing Time from 120 Hours to 3 Hours.
- One of the main challenges the client faced was loading and processing files in an on-premise database in a timely manner. To automate processing to increase efficiency and reduce human error, Tallan used Microsoft Azure tools.
- As a test, one of seventeen external files in the process was combined with on-premise data in an Azure Data lake. Tallan solved the client’s challenge using Azure Databricks, and the process was completed in three hours. The combined data was loaded into Power BI to generate a report.
What’s Large in this Case?.
- Tallan helped this client successfully transform its existing analytics and reporting capabilities to allow analysts to study larger external datasets and project optimal outcomes. The size of the data run used in the pilot was an incoming file consisting of 120 million rows, while the output file generated 520 million rows, for context on what ‘large’ meant in this use-case.
- A centralized data solution has helped improve governance, minimize data fragmentation, and increase speed and efficacy, allowing this firm to serve its customers better.