Tecton, the enterprise feature store that manages the complete lifecycle of Machine Learning (ML) features, and Feast, the popular open source feature store, are now both natively integrated with the Snowflake Data Cloud
Data teams can use Tecton and Feast to build ML features on Snowflake in minutes and serve them to production
SAN FRANCISCO, March 23, 2022 (GLOBE NEWSWIRE) — Tecton, the enterprise feature store company, today announced a partnership with Snowflake, the Data Cloud company, to help data teams operationalize ML applications. The two companies have collaborated to integrate both Tecton and Feast with Snowflake’s Data Cloud. The joint solutions provide a simple and fast path to building production-grade features to support a broad range of operational ML use cases including fraud detection, product recommendations and real-time pricing.
“Cloud data platforms such as Snowflake Data Cloud have emerged as a new option to help data science teams build and deploy ML models. If managed well, these platforms can provide a scalable and governed repository of data to support feature engineering, model training and ML operations. They can offer a system of record for feature stores such as Tecton,” said Kevin Petrie, Vice President of Research at Eckerson Group.
The Tecton feature store is a central hub for ML features, the intelligent data signals that power operational ML models. Tecton allows data teams to define features as code using Python and SQL. Tecton then automates ML data pipelines, generates accurate training datasets and serves features online for real-time inference. With Tecton, data teams can build features collaboratively using software engineering best practices and share features across models and use cases. New features can be delivered in minutes without the need to build bespoke data pipelines.
Snowflake customers can now use Tecton as the interface between the Snowflake Data Cloud and their ML models. Tecton connects to Snowflake as the central source of truth for data, computes data transformations and training datasets on Snowflake using SQL and Snowpark and serves data directly from Snowflake tables.
Simultaneously, Tecton customers benefit from the speed, scalability and cost-effectiveness of the Snowflake elastic performance engine. With Snowflake, ML data pipelines that previously took hours to process can now be delivered in minutes at a fraction of the cost.
In addition, Tecton and Snowflake have jointly built a Snowflake connector for Feast, the popular open source feature store which counts thousands of active users. Snowflake customers can now easily adopt Feast to operationalize analytic data for model training and online inference.
“Many Tecton and Feast users have adopted Snowflake as the central source of truth for their analytical data, and want to use Snowflake as the processing engine for ML features,” said Mike Del Balso, co-founder and CEO of Tecton. “Our Snowflake partnership allows Tecton and Feast users to process ML data on their Snowflake Data Cloud, with best-in-class efficiency and reliability.”
“We are delighted to partner with Tecton to bring commercial and open source feature stores to Snowflake’s platform,” said Torsten Grabs, Director of Product Management at Snowflake. “Snowflake customers now have the option to use either Tecton or Feast to build and manage their features for operational ML.”
For more information on Snowflake and Tecton’s partnership, please visit: https://www.tecton.ai/blog/
About Tecton
Tecton’s mission is to make world-class ML accessible to every company. Tecton enables data scientists to turn raw data into production-ready features, the predictive signals that feed ML models. The founders created the Uber Michelangelo ML platform, and the team has extensive experience building data systems for industry leaders like Google, Facebook, Airbnb and Uber. Tecton is backed by Andreessen Horowitz and Sequoia. The company is headquartered in San Francisco with an office in New York. For more information, visit https://www.tecton.ai or follow @tectonAI.