Snowflake to accelerate ML projects with feature stores from Tecton and Feast

Snowflake is getting new feature stores as more enterprise teams turn to the data company to develop and deploy machine learning applications.

In a statement Wednesday, Tecton announced a partnership with the data giant that will see the former’s feature store, known for managing the full lifecycle of machine learning capabilities, as well as Feast’s open-source store, on the Snowflake Data Cloud to get integrated . As the companies explained, the move will give enterprise data scientists a fast, yet easy way to build production-ready capabilities for a wide range of operational ML use cases, ranging from fraud detection and product recommendation to real-time price tracking.

Problems with ML projects

Companies using cloud data platforms (like Snowflake) for ML projects may encounter issues such as different pipelines during implementation or leaks/inaccuracies in training data. This can slow down development time and affect project delivery. There is also no way for business users to discover and reuse functionality that other data scientists have already created.

With the integration of Tecton and Feast, Snowflake users can easily overcome these challenges. Tecton, a central hub for ML functions, enables data teams to define functions as code using Python and SQL, then automate production-quality ML data pipelines, generate accurate training datasets, and deploy functions online for real-time inference.

“It allows companies to “just turn on” real-time capabilities when they want to get into real-time ML. And it provides a centralized catalog of capabilities for other teams to discover and reuse, greatly accelerating development as ML organizations mature,” Mike Del Balso, Tecton co-founder and CEO, told Venturebeat.

Similarly, Feast – accessible via a Snowflake connector – also acts as an interface to operationalize analytical data for model training and online inference. The open source feature store has thousands of active users and is already integrated with Redshift, BigQuery, Databricks and S3. Meanwhile, Tecton includes companies like Atlassian, Tide, and Fortune 500 insurance companies in its paying customer base and is only integrated with Databricks and S3.

Snowflake’s data science game

The move is Snowflake’s latest step in strengthening its data science operations – one of the six workloads it supports across the data cloud along with data lake, data warehouse, data engineering, data application and data sharing.

“Together with our partners, Snowflake will continue to innovate to improve the user experience at every step of the machine learning workflow, including feature engineering, model training, and model deployment. Snowpark for Python, currently in private preview, is a big step forward. It enables teams to collaborate on a single copy of managed data, use their preferred programming language, and access a rich ecosystem of open source libraries while taking advantage of Snowflake’s Elastic Performance Engine,” said Julian Forero, Senior Product Marketing Manager at Snowflake Venturebeat.

Earlier, the company announced the acquisition of Streamlit to strengthen the data application side of its platform. The deal reportedly closed at $800 million. Snowflake to accelerate ML projects with feature stores from Tecton and Feast

Chris Barrese

InternetCloning is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – The content will be deleted within 24 hours.

Related Articles

Back to top button