Overview

Save up to 60% on Databricks and cloud costs. Zipher is a zero-touch platform that uses ML to automatically optimize Databricks workloads for cost reduction and performance enhancement.

Zipher supports the full suite of SKUs - Jobs, All Purpose clusters, DLT Pipelines and SQL warehouses.

The platform builds profiles of each workload’s behavior by continuously evaluating seasonality patterns and runtime load. Then, it dynamically allocates the optimal resources for every workload and every run, based on available capacity. Zipher’s Spark-aware ML autoscaler understands workload behavior, providing shorter and cheaper runs. Additionally, Zipher enables safe use of spot instances in production by continuously monitoring spot instance availability and autonomously assigning the most available and suitable instances.