Case Study

Challenge: DFC required a modernization of siloed data management systems and financial models to improve the execution of a federal credit program in accordance with OMB standards and Federal Accounting Standards. Past data management and process automation efforts within DFC focused on the development of small-scale tools that could assist in streamlining tasks, often in isolation from connected, downstream tasks. The next generation of automation in DFC needed a strategic approach to implement data management and validation, model deployment, and reporting tools that streamline processes by facilitating the flow of data and calculations from one stage to the next.

Summit's solution: Our team of data scientists, software engineers, and federal finance SMEs built and deployed an Azure-based serverless application to support DFC’s obligation models (finance and funds) and reporting functions. Using Azure DevOps for version control and continuous integration and continuous deployment pipelines, we implemented the development and production environments with a ReactJS user interface, Python-based Azure functions for obligation models, a Python API via API Management Service, Azure Blob Storage for model outputs, and a PostgreSQL database.

Result: This solution allows DFC staff and leadership to conduct critical federal finance business functions through a secure web-accessible application that stores historical model runs and associated metadata. Integrating the two obligation models and ten associated product types into a cloud environment provided numerous benefits, including reduced run-times, fewer model run errors, improved user experience, improved quality control, and increased traceability.