Valohai

Valohai is an end-to-end MLOps platform designed to operationalize machine learning workflows with reproducible, automated pipelines. It provides a central environment where data scientists, ML engineers, and DevOps teams can define, execute, and track experiments at scale across on-premises infrastructure or any cloud provider. Users describe pipelines as code, enabling version-controlled, parameterized workflows that can be triggered manually, on schedule, or via CI/CD systems.

The platform automatically captures metadata for every run, including code version, parameters, data inputs, environment configuration, and outputs, ensuring full reproducibility and auditability. Valohai supports containerized execution, making it straightforward to run training jobs, data preprocessing, hyperparameter tuning, and batch inference on heterogeneous compute resources such as GPUs and distributed clusters.