HOPSWORKS

The collaborative ML platform for batch and real-time data

Easily build and operate feature pipelines, training pipelines and inference pipelines.

View of Hopsworks UI

You design the ML pipelines, we make them work.

Transition from monolithic ML workflows to independant feature pipelines, training pipelines, and inference pipelines. This will enable improved collaboration between your Data and ML teams, and improve operations and maintanability of your ML systems.

For all teams and all data

Increase team productivity and deploy your models faster.

Python Logo

Python-Centric

Feature engineering at reasonable scale. Bring your own code with you, use any popular library and framework in Hopsworks.

Learn more

Collaborative

Role-based access control, project-based multi-tenancy, custom metadata for governance.

Learn more

Spark/SQL/Flink

Feature Engineering at scale, and with the freshest features. Batch or Streaming feature pipelines.

Learn more

All Clouds

Your cloud, your infrastructure, on-premise or anywhere else; managed clusters on AWS, Azure, or GCP.

Learn more

High Performance

Use Python, Spark or Flink with the highest performance pipelines for reading and writing features.

Learn more

Enterprise Support

Enterprise Support available 24/7 on your preferred communication channel. SLOs for your feature store.

Join Slack
Richard Woolston

"Our journey with Hopsworks has been an amazing transformation that's really enabled us to be innovative and reach a point that we wouldn't have been able to reach otherwise.”

Richard Woolston

Data Science Manager - AFCU

ROI with a
Feature Store

Achieve an 80% reduction in cost over time starting from the second ML models are deployed in production.

Read more
Python Logo

Generate value
with your AI Data

MLOps with a feature store allows your organisation to put your data into production, faster.

Read more
Python Logo

Choose the right
Feature Store

Accelerate your machine learning projects and unlock the full potential of your data with our feature store comparison guide.

Read more

Jump right in

Our Python SDK is designed for ease of use. You simply import the Hopsworks library in your Python or PySpark program, connect to your project and start writing data to the feature store.

It is just as easy to create training data from selected features or get inference data for batch or online applications. 

Read our docs to find out more.

# connect to Hopsworks

import hopsworks

project = hopsworks.login()
fs = project.get_feature_store()
# create a feature group

transactions_dataframe = ...

transaction_fg
= fs.get_or_create_feature_group(    
   name="Transactions_30min_aggregates",
   version="1",
   description="Transaction fatures 30 mins window",
   primary_key=['cc_id'],
   event_time='datetime'
)

transaction_fg.insert(transactions_dataframe)

Partners

Community

© Hopsworks 2023. All rights reserved. Various trademarks held by their respective owners.

Privacy Policy
Cookie Policy
Terms and Conditions