Case Study: Overstock.com achieves 5x faster model deployment and nearly 50% lower production costs with Databricks

Overstock.com - Customer Case Study

Overstock.com, a leading online destination for furniture and home décor, struggled to turn billions of annual page views into sales. The company faced heavy ETL workloads that drained scarce engineering resources, data scientists spending too much time on DevOps instead of modeling, and collaboration challenges from teams using different programming languages.

Databricks simplified and sped up ETL, unified multi-language development in shared notebooks, and closed the gap from development to production. As a result, Overstock cut model-to-production costs by nearly 50%, stood up new models five times faster, made inter-day model improvements without redeploys, and used self-service clusters and in-notebook version control to deliver timely, actionable insights.

Overstock.com

Mary Clair Thompson

Data Scientist