At its annual Snowday conference on Wednesday, cloud-based data warehouse company Snowflake unveiled new development interfaces and machine learning features for its Snowpark developer framework.
The new tools, which are mostly in preview, include Snowflake Notebooks, Snowpark ML Modelling API, and Snowpark ML Operations enhancements.
Snowflake Notebooks is a set of new development interfaces that offers an interactive, cell-based programming environment compatible with Python and SQL.
“Snowflake’s built-in notebooks allow developers to write and execute code, train and deploy models using Snowpark ML, visualize results with Streamlit chart elements among other use cases,” the company said.
Snowpark’s ML Modelling API, which is expected to be in general availability soon, allows developers and data scientists to scale out feature engineering and simplify model training by implementing AI and ML frameworks natively on data in Snowflake, it added.
Separately, Snowflake is adding new features to Snowpark ML Operations, which includes an update to Snowpark Model Registry and an integrated Snowflake Feature Store, which creates, stores, manages, and serves machine learning features for model training and inference.
The Feature Store is currently in private preview.
In order to further automate development operations for applications and data pipelines, the company is adding a new feature, dubbed Database Change Management, which is expected to be in private preview soon.
“Using this feature developers can code declaratively and easily templatize their work to manage Snowflake objects across multiple environments,” the company said.
Other updates includes the release of its Native Application Framework on major cloud service providers. While the Native Application Framework is expected to be made generally available on AWS soon, it is expected to enter public preview for Microsoft Azure.
Snowflake plans cost management feature
In order to help enterprises manage their expenditure on Snowflake, the company is planning a new cost management interface, currently in private preview.
The new interface, according to the company, brings together existing cost management features and adds new capabilities so that administrators can go to a single place to get more visibility into account-level usage and spend metrics.
The cost management interface will also show enterprises how the effective value of their Snowflake credits are changing over time due to Snowflake’s continued performance improvements, the company said, adding that enterprises can set limits and notifications to control spending, and can optimize their resource allocation on Snowflake through recommendations.
The recommendations feature is expected to be in private preview soon.
Snowflake, like most other technology providers, has been under constant pressure to help enterprises reduce optimize expenditures, analysts said.
“Snowflake has been fighting complaints around rapidly increasing costs as cloud data warehouses rapidly scale,” said Hyoun Park, principal analyst at Amalgam Insights.
The new interface, according to Park, is part of an effort to counter these concerns and retain large data warehouses where customers may consider migrating to other solutions.
The weight of the pressure for cost optimization on the company can be gauged with the presence of existing third-party, cost-management applications such as Bluesky and Finout, Park said.
Copyright © 2023 IDG Communications, Inc.
