The Snowflake Summit 2023 had several significant announcements regarding new features and improvements to the Snowflake platform. Here’s a summary of the key takeaways for data analysts:
- Dynamic Tables (Preview): A new feature that simplifies data engineering in Snowflake by automating data transformation pipelines. Instead of manually defining data transformation steps and monitoring dependencies and scheduling, users can define the desired end state using dynamic tables and leave the pipeline management to Snowflake.
- Amazon S3-compatible Storage (General Availability): Snowflake now supports accessing data in Amazon S3-compatible storage. This feature allows users to create external stages for on-premises or other cloud storage services and devices compliant with the Amazon S3 REST API, making it easier to manage, govern, and analyze data regardless of its storage location.
- Passing References for Tables, Views, Functions, and Queries to Stored Procedures (Preview): Users can now pass references to tables, views, functions, and queries to stored procedures. The stored procedure performs actions using the active role or secondary roles of the user who created the reference, enabling more flexible and granular control over data access and manipulation.
- Snowpark ML: Machine Learning at Scale (Preview): A new set of Python tools, including SDKs and infrastructure, for building and deploying machine learning models within Snowflake. Snowpark ML supports preprocessing and modeling classes from popular machine learning libraries such as scikit-learn, xgboost, and lightgbm, and works with Snowpark Python for training and prediction.
- ML-Powered Functions (Preview): Three new analysis tools powered by machine learning algorithms:
- Forecasting: Predicts future metric values based on historical data trends.
- Anomaly Detection: Identifies metric values that deviate from typical expectations.
- Contribution Explorer: Helps users find dimensions and values that impact a metric in unexpected ways.
- Native Applications Framework (Preview): Allows users to create data applications that extend the capabilities of other Snowflake features by sharing data and business logic with other Snowflake accounts.
- Custom Event Billing for Applications (Preview): Introduces a usage-based pricing plan that providers can use to charge consumers for usage of apps built with the Snowflake Native Apps Framework.
- Marketplace Capacity Drawdown Program (General Availability): Enables eligible customers with a Capacity contract at Snowflake to pay for listings with their committed Capacity.
For more detailed information and guidance on implementing these features in your Snowflake environment, reach out to The Data Driven Co. As experts in Snowflake, The Data Driven Co can help you navigate these updates and leverage the full potential of your data platform. Contact us today to explore how we can support your data analysis and transformation needs.