In the searing heat of Las Vegas this summer, more than 9,000 customers, users and partners gathered for the Snowflake Summit 2022. DTSQUARED, as a Select Services Partner, was there to hear the keynote speeches live and learn first-hand of the Snowflake innovations and improvements.
When Snowflake launched, its founders set out to build a platform that would allow their customers to fully harness the power of the cloud. Today, the innovative Snowflake Data Cloud offers companies the ability to store, organize, analyse and share vast quantities of data. Clients migrating from other data platforms to Snowflake can run all their data workloads on a single platform, access its unique features and expect to enjoy a financial payback on their investment within months.
Here we discuss the key points DTSQUARED brought home from the Snowflake Summit, as announced by Frank Slootman (Chairman and CEO) and Christian Kleinerman (SVP of Product) and demonstrated by many of their engineers.
Let us start with Unistore (made possible by Hybrid Tables), something Snowflake is hugely excited about. Databases and data warehouses support similar sounding capabilities, but at their core they are very different (transactions vs. analytics, OLTP vs. OLAP, row store vs. column store) and so two different product types have emerged. The launch of Unistore in Snowflake means both use cases can now be supported in one single platform. We eagerly await the opportunity to try this capability and have a great Financial Services use case in mind as a proving ground. The initial offering demonstrated looks very well formed already, so time will tell if this new Snowflake capability may pose a threat to the some of the biggest and longest established database-only companies.
Next came the announcement of Data Apps, leveraging Snowflake’s recent acquisition of Streamlit. Historically, a number of technologies are required to read and write data from the Data Warehouse, leverage Machine Learning, and present results to business users. But now that Snowflake has an embedded app capability, a single platform can do everything. Data no longer moves to the apps, but instead the apps have moved to the data. At DTSQUARED we are already developing a front-end to our bespoke data transformation tool in Streamlit, and as more features are added to Streamlit the potential capability of Data Apps will be incredible.
Other Snowflake announcements included support for Iceberg Tables, currently in development, which will allow users to work with Apache Iceberg (a popular open table format) on external storage while taking advantage of the ease-of-use, performance, and consistent governance of the Snowflake platform, simplifying overall Data Management and enabling architectural flexibility. There was also the enhanced Python Support in the Data Cloud with Snowpark, allowing data engineers and data scientists to code data pipelines and deliver machine learning (ML) workflows. Given existing Snowpark support for Java and Scala, engineers can now use their language of choice to deliver value to customers.
Snowflake’s Cybersecurity workload offers a solution to address the shortcomings of legacy security information and event management solutions (SIEMs). The Data Cloud eliminates data silos, data ingestion and retention limits, and concerns with expensive costs or scalability issues. These and other barriers have restricted security teams from obtaining the highest value from their security data. With Snowflake, security teams can now unify their data, deliver high-fidelity threat detection, and respond quickly to incidents. Splunk had their annual conference in Vegas in the same week as Snowflake; perhaps they found time to check out the emerging competition.
The Data Mesh philosophy is that federated data teams will address the bottlenecks inherent in centralised teams. This can be a huge step forward, as separating your data teams by domain allows teams to work independently but with a common overarching capability sitting over them. A Centre of Excellence typically sets these common standards and provides best of breed technology platforms, such as the DataOps.LIVE CICD platform, to ensure a consistent and efficient approach is adopted in each domain. The many talks on this topic showed how this theoretical construct has now become an efficient and effective physical reality and that Snowflake makes for the perfect platform to act as its core.
As well as presentations in the vast conference halls, there were hundreds of vendors exhibiting at the Summit. Many of their offerings are either Snowflake-only or are clearly building out their new capabilities on Snowflake before they support competing vendors, a clear endorsement of the direction of travel in industry towards the Snowflake Data Cloud.
The United States clearly remains the leader in terms of Snowflake adoption and is a hotbed for vendor innovation too. Roughly half of the world’s top banks have already signed up to Snowflake and while many are at the start of their journey, some – like Capital One – have many years of experience and have even now begun to release their own Snowflake-centric tooling. Capital One ‘Slingshot’ for example is a platform to review and optimise Snowflake credit consumption. Our customers in the UK and Continental Europe are catching up to their US peers, and others not yet on the journey may need to make a start very soon as their slow-to-change legacy platforms get ever harder and more expensive to support.
If you thought this year’s Snowflake Summit was huge, just consider the potential scale of future events based on revenue forecasts. Snowflake recorded revenues of $1.2 Billion last year and already has a further $2.6 Billion of revenue booked but not yet consumed. Their target, by 2029, is to hit $10 Billion in revenue per annum. If achieved, it’s hard to imagine that even Vegas will have a conference centre big enough to accommodate the sheer numbers who may wish to attend. We suggest that you book early for next year to avoid disappointment!
Get in touch with us today to discuss the solutions DTSQUARED can offer to help you get the most out of your data.