Case Study: General Fusion achieves reproducible, version-controlled, and scalable data workflows with Pachyderm

General Fusion - Customer Case Study

General Fusion, a company developing commercially viable fusion power, had amassed large, complex datasets from thousands of sensors in its plasma injector experiments and had outgrown its existing data infrastructure. Needing to augment (not replace) current workflows and enable ad‑hoc sharing with external scientific partners, General Fusion evaluated options and selected Pachyderm for its version‑control features and ability to scale and manage unstructured experimental data.

Pachyderm delivered a language‑agnostic, version‑controlled data platform that treats data like code (repositories, commits, diffs) and tracks full data lineage so every result is reproducible and verifiable. As a result, General Fusion researchers stopped copying data locally, avoided hidden calibration changes, achieved 100% reproducibility of analyses, and can scale and share large experimental datasets more easily — with Pachyderm’s team also providing deployment and support services.

General Fusion

Jonathan Fraser

Engineer