Using feature store

Dear all,

I am going to setup a machine learning pipeline for my project and for the feature store section, I am going to use Hopsworks feature store (managed by myself). According to the Hopsworks’ website, there are two options available, Community and Enterprise editions. I want to have access to all feature store’s aspects such as online features and distributed deep learning on GPUs and it is very important for us that the feature store be scalable. Unfortunately, I could not understand the Community edition of Hopsworks feature store cover my requirements or not. Also, it is not clear for me that the Hopsworks feature store (with enterprise edition features) is free to use or I should pay for a license to use the feature store on my infrastructure.

Hi @mousavihasans,

In regards to the feature store, Hopswoks Community does not include connectors to Sagemaker and Databricks but you can use the online feature store and distributed deep learning with GPUs.

More details on Hopsworks Enterprise features and how to obtain licenses and/or support/training, is available at https://www.logicalclocks.com/pricing

Thanks for your reply

It’s still not clear for me, I am wondering if you can guide me to solve my problem. I am going to restate my question. Is the enterprise edition’s cost just for hardware and servers that are provided by Hopsworks? In the other words, if I use Hopsworks Github source code on my hardware (including clusters, GPUs and etc), can I achieve the same result as the Enterprise edition?

In regards to the online feature store and distributed deep learning you can achieve the same results by installing the GitHub version.