Hopswork vs Feast


I am looking for an open-source solution for the ML platform and meet Hopswork. I tried it on AWS and it looks great after I run the mnist example in Jupyter Notebook and be able to serve the model with a few clicks. Great work!

From what I know, Feast is also open-source, and recently started its cooperation with Tecton.ai. I would love to see a feature-wise comparison between Hopswork and Feast to help me make the choice. My experiment will hopefully lead to the adoption in my team.

If you have done the comparison before, could you please share and let us know? Thank you!

There is a comparison table on featurestore.org, but i don’t think it is appropriate for us to compare ourselves with Feast here. Feast is really a data layer (no UI, no feature computation). Hopsworks is more than that - it is the data layer, feature computation, UI, model training, and serving.

1 Like