Sports team management service revamped for better insights

Spond's rapid growth resulted in significant technical debt. Their data architecture consisted of scattered, fragile pipelines using a mix of disparate technologies like AWS Glue, Lambda, and Python on K8, all feeding into S3 then fed into Databricks. This complex system was taxing on the operational application layer, lacked a single source of truth, was difficult to maintain, and required extensive operational maintenance, preventing the data team from focusing on generating business insights.

We engineered a unified and streamlined data platform centered on Databricks. By consolidating ingestion paths and leveraging Databricks Workflows for orchestration, we eliminated redundant services, which drastically cut infrastructure costs and simplified maintenance. We helped migrate their BI from Looker to Omni, tightly integrated with their new dbt semantic layer. The overall architecture was simplified from a fragmented, multi-pipeline system to a consolidated platform on Databricks. The unified platform leverages Databricks, dbt, and Fivetran for streamlined data flow. Democratized data access now enables a more flexible, self-serve analytics experience across the company.