Snowflake product VP expounds on data warehouse for the cloud

Snowflake, a vendor of data warehouse for the cloud technology, is on the move.

The 2012 startup has seen dramatic financial growth over the past few years. And at its inaugural user conference in June, it showcased a series of new capabilities in the works for its customers.

Christian Kleinerman, vice president of product management, joined Snowflake in February from Google, where he was director of product management at YouTube. And it's Kleinerman who is responsible for the growth and innovation of the company's products.

According to Snowflake, it tripled both its revenue and customer base during its fiscal year ended Jan. 31, 2019. In addition, it raised $450 million in venture capital in October 2018 to bring its total financing to $920 million and overall valuation to $3.9 billion.

Meanwhile, in a move to build on the increasing popularity of data warehouse for the cloud technology, Snowflake on May 1 brought in Frank Slootman as its new chief executive officer. Slootman was chairman and CEO of ServiceNow from 2011 to 2017 and guided it through a successful initial public offering and growth from $100 million in revenue to $1.4 billion. Before that, he was chairman and CEO of Data Domain, guiding that company through an IPO, as well.

In this Q&A, Kleinerman discusses what is taking place at Snowflake and the company's plans for the future.

What service does Snowflake provide, and how does involve business intelligence?

Christian Kleinerman

Christian Kleinerman: Snowflake is a data warehouse and data analytics platform. And, basically, it does what you would expect from a traditional relational database focusing on analytics. But, of course, we have been expanding use cases and adjacent capabilities based on customer feedback and based on the success of that we're having. But it's a core analytics platform.

How does Snowflake's recent partnership with Google Cloud enhance your capabilities as a data warehouse for the cloud?

Kleinerman: We think of Snowflake's footprint as aspiring to be in as many clouds and regions as possible, and our roadmap is enhanced based on customer demand. We're expanding both regions within Amazon and Microsoft, as well as the entrance into Google Cloud, entirely based on customer demand.

We're meeting customers where they are. So, wherever the rest of their applications or infrastructure are, that's where we want to have Snowflake present. And Google Cloud Platform (GCP) just provides more choice for our customers, and choice is always good.

Are there more partnerships that are on the immediate horizon?

Kleinerman: Nothing at this point. We are constantly hearing about both customer demand, as well as partners approaching us. But, right now, we're focused on making the GCP [partnership] a reality and, of course, continue to expand into new regions on Amazon and Microsoft.

What are some of the challenges Snowflake has encountered with migrating on-premises users to the cloud?

Kleinerman: It depends on what the migrating system is and how much time they have with the system, which sometimes translates into legacy. For example, for moving from an on-prem relational database with little business logic, the migration is fairly straightforward. You can just move your data, and you can start running queries fairly easily.

There are two factors that have led to complications. One of them is around integration with third-party tools and systems that sometimes nobody in the company knows much about. We've been requested to integrate with legacy recording BI tools that sometimes people within a company don't even know who's using them and how they were created, but they need them to enable them to complete the migration. So, legacy enterprise infrastructure is an interesting migration blocker. Sometimes customers decide to modernize along the way.

The other point that sometimes adds friction is the notion of business logic. We've encountered customers with millions of lines of code, and understanding those scripts and what that logic is doing is sometimes difficult for the customers themselves.

What is the roadmap for Snowflake over the next three to five years?

Kleinerman: We want to be the core platform that enables analytics and data experiences. The details on what all comes together in that time frame are fuzzy, but we know that we want to have a global presence -- we want to be multi-cloud and multi-region, and make sure to provide choice to customers. We want to enable easy transformation of data ingestion and transformation of data into Snowflake as easy ways get insight out of data and basically convert data into value.

How about the roadmap for the next year?

Kleinerman: We announced tons of new capabilities -- or tons of new directions -- at our summit conference four weeks ago.

We announced the upcoming availability within the next year of our Database Replication capability where we are enabling transactional database replication from any one Snowflake account to any other Snowflake account, whether they are in the same region or the same cloud or not. We're effectively freeing the data and making it easy for any company that wants to pursue a multi-region, multi-cloud strategy, to make it almost frictionless, or largely frictionless.

Another capability that we announced is Business Continuity, where we will deliver failover and failback of a running Snowflake interface, and we will allow client applications to transparently redirect to whatever is the active account.

Another interesting capability that we announced is our Data Exchange, where basically we're delivering a marketplace for our customers to make data available to be consumed by other customers all via Snowflake data sharing, which is something we've had on the market for close to two years.

One other area which is not a focused investment but we're seeing customers leverage Snowflake is the notion of a data lake. If you think of a data lake, it's just a pattern; it's not a specific technology. And what we've found is that customers are leveraging Snowflake to be the data lake more and more.

Editor's note: This interview has been edited for clarity and conciseness.