Scaling the AI ​​factory through conversational analytics

It’s hard to ignore that AI adoption is growing as many teams lack the repeatable architecture to support it. The answer lies in pairing controlled, curated data with conversational analytics that turn dashboards into decision engines, according to Drew Clarke (pictured, left), senior vice president of product and technology at Qlik Technologies Inc. As a member of Qlik’s senior advisory board, Juan Hurtado (right), Inger’s vice president of business and Analytics Inc. Qlik to focus on its strengths – and then put that focus to work within the most demanding business in the industrial sector.

“We’ll direct the business analyst. We’ll direct the data engineer, build tools for your industry,” Clarke told CUBE. “[Hurtado] he gave us that advice as part of the advisory board, and we took it to heart.”

Clarke and Hurtado spoke with Rob Strechay at Qlik Connect 2026, during an exclusive broadcast on CUBE, SiliconANGLE Media’s live streaming studio. They discussed the statistics of the conversation, the design pattern of the AI ​​factory and Ingersoll Rand’s approach to scaling AI in the discovery-driven business. (*Disclosure below.)

Conversational analytics and transition from dashboards to decisions

Moving beyond dashboards to decision automation is exactly what organizations like Ingersoll Rand are pursuing. Rather than rebuilding data pipelines from scratch for each new AI step, Ingersoll Rand focused its approach on the gravity of data already embedded in its Qlik business intelligence platform, Hurtado noted.

“We started to look at, ‘How do we make the right agent conversation statistics from the data we’ve already selected?'” Hurtado said. “We didn’t need all these data engineering efforts.”

The practical benefit is building an “AI factory” — a modular, standardized pattern that respects the semantic layers built into Qlik, handles both structured and unstructured data, and avoids point-by-point solutions that accumulate technical debt. Since Ingersoll Rand completes one acquisition per month, that duplication is not an option, Hurtado explained. Instead, it is the only way to enter new companies without rebuilding the data center each time.

“Those architectural patterns and the AI ​​factory help us do something that has been preached for a long time,” he said. “You do it [proofs of concept] quickly, quickly prototype and reach business grade. “

The need for collaboration is captured by a term the team is now using internally: “the Goldilocks environment” – a lakehouse architecture built on open standards like Apache Iceberg where data stays where it’s needed, always accessible without too much duplication and back to the analytics layer just right, according to Clarke. Buildings are already in production at Ingersoll Rand, he added.

“We want to work together. Don’t create a monolith of everything,” Clarke said. “Have the data where it needs to be, make it accessible, return it in the right way – like Goldilocks.”

Here is the full video interview, part of SiliconANGLE and CUBE for Qlik Connect 2026:

(* Disclosure: TheCUBE is a paid media partner of Qlik Connect. Qlik, the CUBE’s event sponsor, or other sponsors have editorial control over content on TheCUBE or SiliconANGLE.)

Photo: SiliconANGLE