Driving AI-Powered Analytics with Mike Potter of Qlik - Tech Talks

Great to be here.

I think that's a great definition. I also would describe it as my job is to make sure everything works, it scales, it's of high quality, and it solves problems.

Well, big question, a lot to unpack there. Our vision behind Qlik is really to manage the entire lifecycle of data, whether it be from source, getting it into targets, taking it from targets and turning it into true analytic value experiences, AI experiences, and then translating that subsequently into action. This problem has been around for a very long time and has taken many forms. What we've been able to do at Qlik is take that and put a focused effort on making it an integrated solution. More importantly, allowing people to get to their data regardless of where it is and be able to interact with it on their terms, whether it be an on-premise strategy or a cloud strategy. And then finally, being able to do that in a governed way at scale.

I think what we're doing here, particularly as it relates to AI, is being able to integrate it in all phases of the data process. Being able to integrate it into transformations, into how we access data, how we direct what a pipeline would look like, and how we capture information around metadata in our catalog. How we're able to create business glossaries on the fly. There's a lot of feature-centric work that can be automated through the introduction and adding of AI into those processes. All that's really designed to do is take those rote tasks and allow the data engineer, data steward, or data product manager to do their jobs more effectively, really focusing on the content rather than the mechanics. In addition to that, it allows us to create a consistency of approach, so regardless of what data source we're going after and what process we're following, the rules we create with the help of AI allow us to do that in a governed way. I think that's one of the overlooked aspects of AI. People talk about agentic AI and think about autonomous agents ordering lunch for you or whatever. Honestly, one of the most valuable aspects of AI right now is taking unstructured data and deriving data points from that in a structured fashion.

Unstructured and structured data coinciding have always been a problem. I remember lots of iterations of attempts to try to solve it over the years. What we want to do is recognize that they both bring value in their own independent ways and how do you bring them together in a combined fashion. A lot of our AI strategy, particularly on the analytic side, is to introduce the notion of unstructured data into the analytic process through Qlik Answers. You can ask questions that translate into meaningful structured queries and meaningful unstructured queries, and then see how they relate to each other. The biggest opportunity is really to uncover a new way for end users to look at their data other than having to go through traditional user interfaces.

Frankly, one of the things that GenAI has done, particularly with ChatGPT and all the things that have followed, is broaden the horizon of users with less technical skills to be able to get value from the data. That broadening is equally relevant inside of an organization. Heritage BI and heritage analytics solutions have often been the tool of the few that would have to be translated for the many. Now we have the ability to actually allow the many to interact with the data at the same level on an equal footing.

That's right. One of the traps with systems using generative AI approaches is that by nature of the technology, you may get answers that are hallucinations, or where there may be more than one right answer to the same question. What you want to be able to do is provide that information in context with structured data so that the information provides an additional point of view that informs the decision. What it's doing is converging a technology that fundamentally is non-deterministic with algorithmic approaches that are very deterministic, so that you get an overall additive experience.

We've made tremendous strides with respect to our approach to AI up to this point. We just talked about the Agenda framework this week at the conference. We feel really confident in what we've built and we're really looking forward to getting it out there. We have built some very key agents that work very nicely with our analytic engine. The key point we made this week is that even with agentic technology and our addition of AI to the product, our differentiation is still our associative analytic engine, and the fact that these can now work together is really what the win is about. More importantly, it sets us up for the next step where we're going to continue to expand the role of agents within the architecture.

One of the core tenets of what we've built and our architecture is the ability to support integration. Not only integration in terms of who we use as endpoints for data, but also integration programmatically through our open APIs and our platform. We have a lot of customers that have built complete solutions on top of us at a variety of points during the data lifecycle chain to add value. We have a number of those partners here with us this week. They were on screen at the conference, and really they're celebrating the value we bring to them by solving problems that are unique, whether it's a vertical or a particular technology focus. It's really about creating an extendable environment. That goes to our strategic relationships as well. We have very deep relationships with AWS, but we also have relationships with other major players in the space. The whole goal is to make sure that whether it's a data storage technology, AI technology, or cloud technology, we're able to work in the ecosystem that the customer wants us to work in.

I think the argument is not whether they should be in cloud. The argument is how do we get there fast enough. A lot of customers have made tremendous investments in their current on-premise technology, and those investments work and they're very happy with them. What they want to be able to do is take those investments and extend them into the cloud and come up with plans of how to move forward. We've been working with customers in terms of their cloud journey, but there's really no question that cloud is the destination that a lot of people want to be.

We just recently acquired a tool from one of our partners, a great example of a technical integration through our extensible platform that allows improving the accuracy and the volume in which customers can migrate content from their current environments into Qlik Cloud.

Most definitely. We want to partner, absolutely.

Obviously people have access to the content of the conference. They can get access through qlik.com. That is your portal to getting information. You can obviously reach out to us. We'd be more than happy to have a conversation with you. That conversation ranges from what your strategy is for data analytics, what the combined strategy is, and how you get there from here. A lot of customers are on different parts of this journey with different requirements, and we would love to insert ourselves into where you are and help you get from where you are to where you want to be.