Snowflake Inc. today introduced an array of new capabilities for its cloud data warehouse, including a developer tool called Snowpark that will enable companies to deploy custom data wrangling workflows on the platform.

Snowflake went public in a blockbuster September listing that raised close to $4 billion. The company said in the regulatory filing for the listing that it has more than 3,100 customers running more than a half-billion daily queries.

Snowpark, the newly introduced developer tool, allows software engineers to deploy custom code on Snowflake’s data warehouse to perform various information management tasks. There’s initial support for the popular Java, Scala and Python programming languages. 

Snowpark lends itself to creating so-called ETL and ELT workflows for importing records from outside systems into a company’s Snowflake deployment. ETL stands for extract/transform/load and ELT is the process of loading data first and then transforming it later with tools that involve the business users of that data.

Engineers can also create workflows for data preparation, or the process of removing errors from datasets and transforming them into a form that lends itself better to analysis. Yet another use case Snowpark promises to boost is feature engineering, a technique used in artificial intelligence development to convert information into a form that is easier for a neural network to work with.

A second enhancement Snowflake announced today is support for unstructured data. Until now, the company’s data warehouse was mainly geared towards storing structured information organized in rows and columns and semi-structured information, like sensory measurements from connected devices. Now, Snowflake is adding unstructured records such as audio files, video, PDF documents and images to the list of supported formats.

The enhancement will enable Snowflake’s enterprise customers to store more of their information in its platform. That could have a positive impact on the company’s revenues because it provides its data warehouse under a usage-based pricing model.

To help customers secure the datasets they keep in its platform, Snowflake today introduced the ability to set row-based data access restrictions. That means organizations can configure the data warehouse to let users retrieve only the specific rows of information they need for their work. Such restrictions are important for managing datasets containing details of varying sensitivity, for example customer datasets that include both purchase histories and credit card numbers.

Snowpark is available today in Snowflake customers’ test environments. The unstructured data support, meanwhile, is currently in private preview, while the row-based access controls are expected to enter private preview later this year. 

Image: Snowflake