Data science stack (DSS)

Data science stack (DSS) is a ready-to-run environment for machine learning and data science on Linux. It combines open-source tooling, including MicroK8s, JupyterLab, and MLFlow, into a setup intended for work on Ubuntu or other Snap-enabled workstations. DSS is distributed under an open-source license.

The product provides a command-line interface for managing containerised machine learning environment images. These images can include environments based on PyTorch or TensorFlow and run on top of MicroK8s. By using isolated and reproducible environments, DSS addresses the configuration work normally required to prepare a workstation for machine learning. It is designed to use the workstation’s GPUs when those resources are available.

DSS is intended for both people beginning with machine learning and engineers who need to create more complex development or runtime environments. Its approach reduces the amount of manual setup required and helps users start working without building each environment from scratch. The same environment model can support development work and production-oriented preparation.

In the MirrorNest download catalog, DSS is listed in the Developer Tools category for the Linux platform. MirrorNest mirrors the official build directly onto its own storage rather than presenting an independently modified package. The mirrored build is scanned for malware before publishing, giving readers a clear distribution path for obtaining the open-source Linux tool.