Use Cases

Leverage the power of Git at 100x the scale.
XetHub accelerates big data workflows on your desktop and in the cloud.

Model training and development

ML development doesn’t need to wait on data downloads anymore.

Learn more

Data exploration and preparation

Use your tools to interact with big from your desktop.

Learn more

Large file comparison and evolution

Iteration on any file type made safe, fast, and understandable.

Learn more

Model training and development

  • One line of code to access your data, no matter where you're running
  • Mount terabytes of data in seconds to kickstart your distributed training
  • Fearlessly iterate on code and data in one place with always-in-sync development
  • Built-in reproducibility with models stored alongside the code and data that generated them
  • Automatically generate reports with CI/CD and custom visualization
A chart showing git-xet mount speeding up training time; a directory of code, data, and models; and an illustration of a sample regression visualization
An illustration of a custom visualization and summary view

Data exploration and preparation

  • Easily access huge data with local tools using instant mount
  • Understand data at a glance with automatic summary views
  • Confidently change data with built-in version control
  • Show your data in context with custom visualizations
  • Collaboratively review dataset iterations with pull requests
  • Ensure data quality with integrated CI/CD flows

File comparison and evolution at scale

  • Track terabytes of code and assets with familiar Git syntax
  • Conveniently mount multiple versions of large data locally for easy access and comparison using desktop tools
  • Speed development while saving time and bandwidth with difference-only data transfers
  • Store large file iterations efficiently with content-aware block-level deduplication
  • Infinitely rewind and replay your changes, with browseable metadata and visualizations for understandability
An illustration of using git on the command line with large file deduplication