11-09, 16:15–17:00 (America/New_York), Music Box (5th floor)
Version control for data is a critical but underserved component of modern data science. Besides giving you reproducibility, true data version control enables diffs and collaboration. In this talk, we'll introduce Dolt, a free and open source version controlled database modeled after git, and demonstrate how you can use it to add reproducibility and the other benefits of data version control to your data pipeline. This talk is for data scientists and engineers, especially those who write scripts to automate pipelines. Working knowledge of git is very helpful but not required.
Dolt is Git for data: a new kind of SQL database that you can branch and merge, push and pull, fork and clone just like a git repository. Learn how to use Dolt to add reproducibility and a wealth of other version control benefits to your data processing pipeline.
DoltHub is GitHub for data: a place where you can host Dolt databases for free, others can clone and fork them, submit pull requests to improve your data, and bunch of other features.
Dolt is free and open source, and DoltHub is free for public databases.
The talk includes a live demo of using Dolt to manage branching and merging a data set, and as well as showing you how to upload changes to DoltHub and open a pull request for your team to review.
No previous knowledge expected
Zach leads development for Dolt, the world's first SQL database that you can fork and clone, branch and merge, push and pull just like a git repository. Zach studied computer science at the University of Washington, and spent the first 13 years of his career split between Amazon and Google before joining DoltHub. He's a fierce advocate for the value of client-side software in a server-side world.