VESSL enables seamless scaling of containerized ML workloads from a personal laptop to cloud instances or Kubernetes-backed on-premise clusters.

While VESSL comes with an out-of-the-box, fully managed AWS, you can also integrate an unlimited number of (1) personal Linux machines, (2) on-premise GPU servers, and (3) private clouds. You can then use VESSL as a single point of access to multiple clusters.

VESSL Clusters simplifies the end-to-end management of large-scale, organization-wide ML infrastructure from integration to monitoring. These features are available under 🗂️ Clusters.

  • Single-command Integration — Set up a hybrid or multi-cloud infrastructure with a single command.
  • GPU-accelerated workloads — Run training, optimization, and inference tasks on GPUs in seconds
  • Resource optimization — Match and scale workloads automatically based on the required compute resources
  • Cluster Dashboard — Monitor real-time usage and incident & health status of clusters down to each node.
  • Reproducibility — Record runtime metadata such as hardware and instance specifications.