Tacton launches AI assistant to cut CPQ modelling work

Tacton has launched an AI Product Modeling Assistant for its configure, price and quote software, saying it cuts product modelling effort by 50%.

The feature is aimed at manufacturers of complex, configurable products, where product modelling often takes the most time in a CPQ deployment. It turns product information from spreadsheets, technical specifications, PDFs and legacy data into a structured model that is typically 70% to 80% complete before human review.

That matters because product modelling determines how much of a manufacturer's catalogue is available when a CPQ system goes live and how quickly the business can start using the software. It has also become harder for some manufacturers to sustain that work as experienced engineers retire and specialist modelling expertise becomes scarcer.

Manufacturing focus

The assistant is designed specifically for manufacturing rather than as a general AI add-on. According to Tacton, it recognises the structure of configurable products, including assemblies, attributes and constraints, and produces models in an environment where staff must review and approve outputs before use.

Jesper Alfredsson, chief product officer at Tacton, outlined the company's view of how AI should be used in industrial software. "As manufacturers move toward a buyer-centric smart factory, AI needs to be built into the core workflows that drive real outcomes," he said.

"We don't layer generic AI on top of CPQ. Our focus is on applying innovation where it has the greatest impact on timelines, with full governance and control."

The launch reflects a wider effort by software suppliers to apply AI to narrower operational tasks rather than broad conversational tools. In manufacturing software, vendors are targeting areas where engineers and sales teams still rely on manual work to organise product rules, variants and dependencies before orders can be configured accurately.

Customer examples

Tacton cited early users that said they had reduced modelling time while rolling out CPQ across different products and regions. One was Alimak, which sells vertical access systems.

"We've seen between 40 to 80 percent improvement in how quickly we can build product models," said Frank Klessens, group product support manager at Alimak.

"That allows us to bring products to market faster and focus more on refining and improving the model instead of building everything from scratch."

PERI SE, which supplies formwork and scaffolding systems, also reported shorter modelling cycles. Its example suggests the software can reduce not only initial build time but also the delay before internal users can begin testing products in a CPQ environment.

"The Modeling Assistant helped us turn a two-day modelling effort into a two-hour task. It has fundamentally changed how quickly we can structure new products in CPQ," said Tobias Gottstein, senior manager sales solutions at PERI SE."

"By leveraging AI to generate new models, including assemblies, domains and module variants, we reduced modelling time in half, enabling the business to start testing products in CPQ about one week earlier than originally planned."

Implementation pressure

For manufacturers selling highly configurable equipment, CPQ systems sit between engineering data and the sales process. They help sales teams assemble product options, apply pricing rules and produce quotes without creating combinations that cannot be built or delivered. But before that can happen, product structures and rules must be modelled in detail, which can be a slow, specialist task.

Tacton's new assistant is intended to shift that step from a manual build process to one where AI proposes a starting point and modellers refine it. The company says this can increase the amount of modelling work existing teams can handle as product portfolios expand and new configurations are introduced.

The move also reflects pressure on industrial companies to shorten software roll-outs and reduce the effort needed to maintain digital product data over time. As manufacturers add variants for different markets and customer requirements, the burden of updating CPQ models can continue long after the initial deployment is complete.

Tacton, which has operated for more than 26 years, sells software to manufacturers of configurable products and is headquartered in Stockholm and Chicago. Its AI Product Modeling Assistant is embedded in the broader CPQ platform and requires human validation before models are synchronised into the system.