Data Science work starts in a notebook, an environment designed for rapid data exploration. However, notebook tooling is underdeveloped, and it is hard to set software development best practices. In addition, moving code into production requires complete refactoring and testing, creating a bottleneck on the data scientists.

We've developed