Faster

Cut Databricks Spark costs by up to 80%.

Most Spark jobs run on wrong-sized machines with incorrect configs, leading to slow jobs, wasted spend, or both. Under the hood, this is a search problem: we're trying to find the lowest-cost configuration that meets your SLA. Testing every config would be slow and expensive; we use ML, alongside a small number of real samples, to understand the shape of your workload and winnow down the options. We've found three consistent patterns for why job configurations are wrong: The settings that control this, instance type and worker count, are usually chosen when somebody creates the job.