Data Utilization in the Cloud Era

Cloud Data Insights (CDI) had the opportunity to talk with Sean Knapp, Founder and CEO of Ascend; and Tom Weeks, Ascend’s CRO, at the Gartner Data & Analytics Summit. We covered shifts in how businesses think of their data and how Ascend is responding to–and even anticipating–emerging challenges around effective data utilization. Here is our conversation.

Note: This interview was edited and condensed for clarity.

CDI: The Gartner Data & Analytics Summit is the ideal forum for data leaders and software and services providers to take the market’s pulse and get a sense of where the technology is headed. The pandemic caused many organizations to use data differently. Have you seen the market shift? Where have you seen your customers’ attention shift? Are you having different conversations than you had a couple of years ago?

Sean Knapp: There has been a material shift. I think that the macro shift that’s happened is that we are no longer worried about data infrastructure. The cloud data vendors have done their job. It’s fantastic–so we’re no longer worried about how many bits and bytes we store, how fast we can process. That’s important as it’s moved the conversation around what the priorities are and frankly, which group has the focus on data and what that focus is has shifted a fair bit. The market is shifting closer to the application of data as opposed to the storage and processing of data. It is now centering around actual productivity. And it’s at a different size and scale. And time–where people want to be able to create data products in hours, not days, at the keyboard.

So, have we alleviated the pains associated with this one problem? Can we declare victory? We have freed everybody from how much you can store and how much you can process. But as a result, we have 5 to 10 times the number of people who can do things with data and even more people building and deriving residual products on top of those. Now we are getting into the product productivity domain, and how do we help people build more things faster–and safely? And that is a fundamental shift in focus. Approaches that we’re hearing about here around the ecosystem are around data mesh and data fabrics or active vs. passive metadata management.

Sean Knapp: Add into that mix DataOps, another category of many different vendors and strategies. To me, the chaos of that space embodies the growing awareness that there are problems that people are trying to solve. We don’t quite know yet how that is going to take shape and solution looks like. We have a strong position on what we think the solution looks like–it’s automation. I don’t think anybody has found ways to scale people in than with true automation. We’ve seen the ability to apply advanced layers of automation to scale human effort in marketing and with Robotic Process Automation (RPA), and we’ve seen it across industries.

CDI: How long do you think it will take to see if automation is indeed the solution to allowing more people to do more with data, as you said, safely?

Sean Knapp: I recently added a question to a survey we’ve been doing: Do you intend, or do you already have automation in place, or do you intend to invest in automation to increase your team’s productivity? The interesting insights were:

3.5% said they already have automation. That was incredibly low.
88.5% plan on investing, with half of those saying they would invest in the next 12 months.

Tom Weeks: And it would have been interesting if we asked what automation means. We probably would’ve gotten 20 different answers. At a meta-level, I’m sure they’d all be around the same thing: scaling people. (Survey results announcement)

CDI: The “data automation cloud” is a term that Ascend has used. Do you anticipate that the automation will happen in the cloud?

Sean Knapp: So we see that there are many more shades of gray than there used to be between just on-prem and in the cloud. Ascend has always done a dedicated deployment of both the data and control planes–dedicated, as in private, without multi-tenancy, and even in the public clouds. We also have a multi-tenant cloud offering product to provide options. For example, we are seeing hybrid scenarios where someone has a traditional data warehouse or wants their ETL pipelines to run on top of a cloud data lake. We’ve also seen real-time replication run across clouds. Our product can connect across clouds and environments.

CDI: What’s on your technology roadmap to meet this growing need for automation?

Sean Knapp: We work in both mesh and fabric architectures, but our metadata and automation capabilities very much cater to a fabric architecture where our customers have for years now pulled us into more and more advanced fabric style architectures where they’re running distributed data and processing across clouds. Watch for exciting announcements later this year.

CDI: What are your priorities for the near future in terms of business growth or new initiatives?

Tom Weeks: We need to focus even more deeply on the developer. The developer has enormous influence, so one of the shifts we made was to listen more closely to the developers. Our product teams aligned more with what developers need and we do everything we can to support them. We believe if you take care of the developers and give them what they want, they will put new workloads on the platform. What do they want? They want better lives, things that are easier, and companies that listen to them.

CDI: Can you give us an example of a developer-driven feature?

Sean Knapp: Our research shows that 65% of data transformation’s built in SQL, and 32% used Python. About 3% is in Java or SCALA. We try to make them all fairly equitable. People tend to gravitate towards what is the most effective. When building data pipelines, the storage and processing infrastructure determines the tools, which generally drives what language you write. And only at the end of that entire decision-making chain does somebody start to think about the actual data product. That’s a little backwards, but we wanted to create a platform that lets a developer focus on the data product and not the tool chain and language and use automation where we can to lighten the load.

CDI: What do you think makes you able to understand what developers want?

Tom Weeks: We like our customers. Seriously now, it can take a lot of digging when exploring what a customer is trying to accomplish. We spend a lot of time watching them and experiencing them using the product. They have something in mind, but if you take that at face value, sometimes you miss the bigger problem they’re trying to solve and how best to solve it—and figuring out how to do that so that everybody can benefit.

CDI: Thanks for your time and especially your insights on automation in data engineering and what developers want. We look forward to those product announcements you mentioned.