Hey everyone! Link if you don't want to read everything below – https://DAGsHub.com
The past few weeks have been pretty crazy. Most of my friends, the people I know and myself have shifted to working from home due to COVID-19. Everyone I’ve spoken to is learning that this change requires many adaptations, one of which is adopting new workflows for remote teamwork. This is true for any team, but my focus is on data science teams.
I am one of the founders of DAGsHub, which is a platform for data science project management and collaboration. The platform is built specifically for community collaboration – meaning people working from different locations on the same project. Due to the situation, we’ve decided to enable private projects with unlimited collaborators for free, available to everyone for the next 6 months.
We’re hoping that using the platform will improve your remote team workflow. For us, this is a chance to get your feedback and better understand your needs so that we can build the most focused platform for your needs.
A bit more about DAGsHub:
The platform is still in beta, and we’re adding new features from user feedback as quickly as we can. At its core, DAGsHub integrates between experiment tracking and reproducibility. We are building the platform based on, and inspired by, great open source tools like git, DVC, MLflow, Cookiecutter data science and more.
The quickest way to get started is to go through our tutorials:
Tutorial for experiment management on DAGsHub
Tutorial for DVC and reproducibility on DAGsHub
An important note about data – since we use DVC for versioning of data, you don’t need to upload your data to DAGsHub to get the benefits. You can use your existing cloud storage (or just a folder on your PC, an SSH server, Google Drive, or any other type of remote supported by DVC), while DAGsHub stores only metadata and links to it. That means that data security and permissions are handled by your storage provider.