It’s “snow time”: Snowflake World Tour AI advancements mark shift to inclusive data

It’s time for data clouds to enable all users to “slice and dice data any way you would like” as Frank Slootman, CEO of Snowflake has put it.

During its Data Cloud World Tour event at ExCel London this week, Snowflake announced new features to its data software offering alongside several partnership developments, marking the platform’s shift to enabling increasingly democratized data access and analytics capabilities across its users.

The firm is launching Document AI, a new analysis technology that uses AI technology to take unstructured objects and turn them into semi-structured information. Information can be searched and drawn from documents, PDFs and more across multiple databases of information, giving users more digestible and useful information at a faster pace.

It’s hoped the tool will also assist with minimizing factors such as egress charges and governance issues, as the data will not shift for users to gain the insights they need. The launch follows Snowflake’s acquisition of Neeva, a GenAI intelligent search company, back in May this year.

Also announced is the general availability of Snowpark Container Services on the Snowflake Platform. “Removing limits on data functions”, it enables any new or legacy function, application or functional database to deploy and run inside of Snowflake within a containerized setting, using open-source and Kubernetes technology.

More than a dozen customers are now reportedly implementing Snowflake’s Unistore, its cloud solution for handling transactional and analytical data, and Slootman teased in his keynote speech that more announcements are set to come in November for this tool and others.

In partnership with J. P. Morgan, the launch of the Fusion Data Mesh aims to meet J. P. Morgan’s clients where they are, using ML and AI within cloud-native channels to retrieve and deliver investment and securities services data directly to their workflows in Snowflake instances and Python notebooks.

It was also shared that Blue Yonder has now re-platformed all its software modules on a Snowflake data cloud for greater visibility of its supply chain data and the removal of data silos.

Other partnership outcomes including Bupa and Bentley were also shared and, through Snowflake, have been able to create ways to democratize data across their organizations in “a safe and secure way”. Notably, Bentley has been able to create a data dojo literacy program, allowing every employee the opportunity to accelerate their upskilling in data science.

Following ongoing updates to its platform, the firm has reported a 15 percent saving for users on Slowflake costs over the course of the year. Other features have also been delivered, such as budget limit settings and greater utilization visibility to match user requirements. Designed to offer easier access throughout the enterprise, views for both data scientists and non-scientists have also been made available with more simple sharing capabilities.

For anomaly detection, a feature called Contribution Explorer allows users to leverage ML to understand where changes originate from without needing to learn Python coding.

Snowflake additionally shared that now over 10,000 Streamlit apps have been launched on top of its platform. Plus, the firm is said to be making the creation and deployment of language models and chatbots possible for a greater variety of people in businesses.

Frank Slootman, CEO, Snowflake, shared in his keynote speech that the time has come to make data more inclusive: “The simple way to think about it is that 80 percent of the world’s data is unstructured, so it’s really important that we are delivering on the promises of AI and ML. It’s one thing to blow up the silos but the hard part is to really enable all the possible workloads that you want to run against the data. This a complete reversal of what we’ve been doing for the last 50 years – where with data strategy, the data stays put and the work comes to the data, instead of the data coming to the work. That’s what the data cloud is all about.

“There are all kinds of issues; you’re crossing governance parameters, accruing egress charges, with a lot of operational complexity. What we see in the fullness of time is that the vast majority of consumption on Snowflake will not be coming from data scientists, data engineers or data analysts, it will actually come from applications, being used by consumers, end users, business people that don’t consume data – they consume data through function.”