How Snowflake is supercharging data applications across industries (VB Live)

Presented by Snowflake

Across use cases and industries, software teams are building their applications in the cloud. What are the advantages? How does it deliver value to end customers? Listen to this VB Live events for key learnings and more from leaders at top software companies.

Access on demand here for free.

According to McKinsey, 75% of the cloud’s predicted value comes from boosting innovation — and is estimated to create $770 billion in innovation-driven growth. Its transformative power is driving disruption in every industry, with so much of its value coming from the way it enables secure, data-driven business agility, product development, data strategies, intelligent solutions, and breakthroughs in software engineering and platforms.

The latest VB Live event, “Innovation on the Data Cloud,” explored how the cloud is changing the game for companies of every size, with a panel of product and engineering leaders who are driving innovation in their roles. Together, they spoke to how the cloud helps them support customers, building an agile, data-powered platform, machine learning and AI, and more. (See all speakers below.)

Supporting customer experience with the cloud

When CloudZero, a cloud cost intelligence company, first started building in the cloud, they took a completely serverless approach. That means they maximized performance and cost efficiency by only developing, building, and delivering functionality that’s being consumed by customers. To scale the platform, they needed a data warehouse that was going to support that same architecture, allowing them to serve their customers’ needs on demand with a user experience that was extremely performant.

“Whether you’re a super large customer of ours, where we manage millions of dollars of spend a month, or a smaller customer managing a smaller cloud build, you have an equal level of responsiveness from our platform,” said Erik Peterson, Founder & CTO/CISO, CloudZero.

“Snowflake drives that by having multiple warehouses we can tap into to drive different parts or functionalities within our system. That capability was critical to us in terms of how we built out and scaled out our platform.”

Securonix, a security operations and analytics platform, had a data repository silo for security data, but it wasn’t well-integrated into their overall data platform, data strategy, and data architecture. Rather than trying a DIY approach to break the barriers and add security data to the rest of their data, they went with a turnkey secure analytics solution and an enterprise data cloud that they could use in the same way they use any other type of big data needs they have — which allowed all the data to be ingested and used in analytics.

“The idea was to leverage Snowflake in a way that organizations could use what they already had in place for their other big data needs,” said Augusto Barros, VP Cyber Security Evangelist at Securonix. “It wasn’t about Securonix replacing its backend with Snowflake, but we’re coming to our customers and saying, if you want a non-siloed approach, we have a partnership with Snowflake. We call it a bring-your-own-Snowflake solution.”

Agility and scalability

“Internally, Snowflake’s versatility and scalability have done so much to benefit us when achieving that agile design, even when we’re operating with very large customer signal data sets,” Walker said.

For Simon Data, a customer data platform with an integrated marketing cloud, it offers agility of workload, from very large, high throughput queries, all the way down to small, low-latency use cases. They use Snowflake as an ELT and an aggregation tool on the large customer data sets, which helps them scale. They’ve also found that Snowflake can be performant enough to power applications that are latency sensitive, such as real-time messaging campaigns. It allowed them to build an event connector that’s specific to Snowflake, which powers what they call database triggers. It lets them fire off behaviorally targeted messaging in an existing real-time campaign platform from the results of the Snowflake query.

There’s also agility in the time to value, said Matt Walker, Simon Data’s Founder & CTO/CISO. Snowflake’s scalable performance meant that they didn’t need to build out complex layered caching over the top of their our unified contact view, which is a standard diagnostic product at the core of any CDP.

Instead, they clustered events by contact identifier, which allowed them to use only one micro-partition per contact query. That got them the response time needed through user-facing queries, even when they were scanning up to three years of data for a single contact. That meant they got the product to market very quickly, and were able to shift their valuable product development resources elsewhere.

Peterson said being able to have the same platform that drove the backend machine learning pipeline, data ingestion processing, and all of the reporting and daily processing that they do, against billions of billing events and cost activities, system activities that they’re correlating within the system, means avoiding complexity and still getting that highly performant user experience.

Looking to cloud providers like Snowflake or AWS to manage the infrastructure is key for both scaling the existing team, or how quickly they can get new functionality to market, and scale in terms of how delightful an experience they can deliver to their customers. They’re achieving both of those things by having a data platform that allows them to handle the heavy lifting processing, where new data is constantly being ingested and pulled into the platform, but at the same time gives customers almost real-time notifications when spending goes out of control.

“Snowflake gives us huge scalability and long term retention,” Barros said.

Many organizations trim the security data they’ll ingest, as well as the retention period, so they can avoid the capacity issues and the expense. Snowflake is cost effective enough to allow organizations to ingest the data they need for threat detection, and to keep that data for longer, which is a huge advantage in cybersecurity. It can be months, sometimes even years, between the first interaction with a threat and when a company actually discovers it, and not having a history of data then becomes a huge liability.

For more on database design, the wide variety of data processing use cases in the cloud, how Snowflake enables seamless data sharing, integration, and much more, catch up on this VB Live event on demand.

Access on demand here.

You’ll learn best practices to: