Snowflake data cloud adds Python, multi-cloud collaboration
The cloud data vendor released preview updates to its platform to accelerate data queries, better support multi-cloud operations and boost developer productivity.
Cloud data vendor Snowflake on Monday released a series of updates to its data cloud platform that aim to accelerate query performance, boost developer productivity and enable multi-cloud collaboration.
Snowflake -- which went public in 2020 -- has been steadily iterating its data cloud with capabilities designed to make it easier for organization to run data analytics, business intelligence and operations workloads.
In June, the vendor introduced a preview of its Unistore Hybrid Tables for combined analytics and transactional workloads as well as support for the Apache Iceberg data lake table format.
The updates revealed today at the vendor's Snowday 2022 conference, held in-person in San Francisco, include a public preview of the Snowflake query acceleration service that aims to speed up large queries by adding more processing power for the biggest queries.
Meanwhile, the vendor's new Snowgrid capability, also in public preview, is intended to enable collaboration and data governance more easily for users across different cloud providers. After a year in preview, Snowflake also made support for the open source Python programming language generally available in its Snowpark developer service.
During a streamed media session ahead of the conference, Snowflake users discussed their experiences with the Snowflake platform and the new developments. Among the users was Fabio Luzzi, global head of data and analytics at Tapestry, which runs a series of fashion brands, including Coach, Kate Spade, and Stuart Weitzman.
Tapestry has hundreds of data sources powering its enterprise applications -- including customer-facing use cases, such as the company's loyalty program and call center -- that help the business manage supply chain operations.
"Our vision has always been, 'How can we monetize our data assets, and how can we leverage all of it to make a financial impact?'" Luzzi said. "Snowflake is a centerpiece of our data strategy."
Tapestry and the Snowflake data cloud
Luzzi said Tapestry is using Snowflake as the foundation for a data exchange layer that helps the company improve its operations.
Fabio LuzziGlobal head of data and analytics, Tapestry
Among the reasons Tapestry uses Snowflake is that the technology can work across multiple clouds. This capability is now expanded with the Snowgrid services.
"The idea of having powerful data related solutions that are cloud agnostic, so it's not locked into any specific cloud, is exciting for us," Luzzi said.
Veronika Durgin, head of data at Saks Fifth Avenue, said her organization is in a similar situation.
"Saks utilizes Snowflake to power our platform, and it hosts all of our data for secure and easy access," Durgin said. "So because the platform holds all of our raw data in one place, it just helps our internal users easily browse and find all company data."
Saks isn't just using Snowflake as a data consolidation technology but also as part of its data analytics and machine learning efforts. For example, Saks uses its data in Snowflake to enhance its user personalization capabilities and improve customer experience.
Building Python application in the Snowflake data cloud
With its Snowpark service, Snowflake lets users like Northern Trust to build applications with the Snowflake data cloud.
The financial services giant's clients include large, sophisticated institutional investors with complex investment portfolios. Getting a complete view of those portfolios and providing updated valuations and market data is a massive data integration and management challenge, said William Wu, head of quant analytics at Northern Trust.
To address that challenge, Northern Trust is using Snowpark to build a product called the Total Portfolio View (TPV).
"TPV relies on Snowflake as our single source of truth for data," he said.
Northern Trust uses a lot of Python in its operations. Wu said that the general availability of support in Snowflake Snowpark is important.
"With Snowflake support for Python, that enables us to build a fully integrated ecosystem that empowers our data science teams to work with the full wealth of foundational data and deliver solutions very quickly to the market," he said.
Streamlit is coming to Snowflake data cloud
Python support in Snowflake's Snowpark will also soon get a big boost after Snowflake's acquisition of open source machine learning app vendor Streamlit earlier this year.
Streamlit's technology lets developers share Python-based web applications. In early 2023, Snowflake plans to release a preview version of an integration of Streamlit into Snowflake to select customers, said Adrien Treuille, head of Streamlit at Snowflake.
"We get to leverage Snowflake's global performance engine to deploy your apps across the organization," Treuille said. "So click a button -- boom, you get a URL. And now you can share your app with others outside your data team."