Using Databricks? Consider Qubole

Reduce CAPEX and OPEX by atleast 50%  Significantly reduce cloud computing costs with sustainable economics through built-in capabilities such as workload-aware autoscaling, automated lifecycle cluster management, intelligent spot management, and heterogeneous cluster configurations.

Maximize Spot Instance Savings while Minimizing Job Interruptions Allow organizations to optimize the use of Spot instances (AWS or Google), resulting in cost savings of up to 80% on compute costs. Qubole provides policy-based automation of Spot instance management to balance performance, cost, and SLA requirements. Qubole scales underlying compute automatically as needed for bursty data workloads while keeping the cost low

Near-Zero Administration  Qubole automates the installation, configuration, and maintenance of clusters, multiple open-source engines, and purpose-built tools for data exploration, ad-hoc analytics, streaming analytics and machine learning. Organizations realize administrator-to-user ratios of 1:200 or higher and near-zero administration experience.

Execute any ML, AI, Streaming or Analytics Use Case  To realize the faster adoption of the data lake, you need a platform that is open for multiple use cases and personas. Qubole includes an out-of-the-box workbench and notebooks for data scientists, data engineers, data analysts, and administrators. It supports open source frameworks used by every type of data user including Apache Spark, Presto, Hive/Hadoop, and Airflow.