Databricks lowers enterprise compute costs by 20 percent

Nucleus interviewed multiple enterprises that adopted Databricks Lakehouse to support a broader range of analytics use cases and found significant improvements in their compute costs, processing latency, and model performance. Customers reported millions in savings by sunsetting legacy data management systems, 25 percent reduced compute costs, 75 percent faster processing via dynamic resource allocation, over 80 percent lower data storage costs by switching from on-demand nodes to spot instances, and 29 percent improved model accuracy by incorporating hundreds of additional features into the organizations’ real-time streaming pipelines. As data collection and utilization continues to scale over the next 18 months, Nucleus expects Databricks Lakehouse to see accelerated adoption, especially among large enterprises with complex legacy infrastructures.