Whitepapers

OptScale is an open source FinOps and MLOps solution built for ML/AI, Big Data, CI/CD and regular workloads

OptScale is an open source FinOps and ML/AI optimization tool available under the Apache 2.0 license. Also users can get the OptScale solution both for on-premise deployment and as SaaS.

OptScale ML/AI application profiling and performance optimization capabilities

ML/AI task profiling and optimization

ML/AI and Data engineering teams get a tool for tracking and profiling ML/AI model training. OptScale collects inside/outside performance and model-specific metrics, which help give performance and cost optimization tips for ML/AI experiments or production tasks.

ML/AI metrics and KPI tracking and transparency across ML/AI teams

OptScale profiles ML/AI models, gives deep analysis of inside/outside metrics to identify training issues and bottlenecks. OptScale improves ML/AI profiling process by getting optimal performance and helps reach the best outcome for ML/AI experiments.

Dozens of tangible performance improvement recommendations

OptScale performance optimization tips include utilizing Reserved/Spot instances and Saving Plans, rightsizing and instance family migration, Spark executors’ idle state, detecting CPU/IO, IOPS inconsistencies that can be caused by data transformations.

Runsets

OptScale enables ML /AI engineers to run a bunch of training jobs based on a pre-defined budget, different hyperparameters, and hardware (leveraging Reserved/Spot instances) to reveal the best and most efficient outcome for your ML/AI model training.

Spark integration

OptScale supports Spark to make Spark ML/AI task profiling process transparent and more efficient. A set of OptScale recommendations, which are delivered to users after profiling ML/AI models, includes avoiding Spark executors’ idle state.

Minimal cloud cost for ML/AI experiments and development

OptScale in-depth cost analysis and dozens of optimization best practices help minimize cloud costs for ML/AI experiments and development. The tool delivers ML/AI metrics and KPI tracking, providing complete transparency across ML/AI teams.

Supported platforms

About us

Hystax develops OptScale, an MLOps & FinOps open source platform that optimizes performance and IT infrastructure cost by analyzing cloud usage, profiling and instrumentation of applications, ML/AI tasks, and cloud PaaS services, and delivering tangible optimization recommendations. The tool aims to find performance bottlenecks, optimize cloud spend and give a complete picture of utilized cloud resources and their usage details. The platform can be used as a SaaS or deployed from source code; it is optimized for ML/AI teams but works with any workload.

Contacts

Email: [email protected]
Phone: +1 628 251 1280
Address: 1250 Borregas Avenue Sunnyvale, CA 94089

Enter your email to be notified about new and relevant content.

You can unsubscribe from these communications at any time. Privacy Policy