Basware has deployed Kantata's professional services automation platform across its services operation, increasing capacity and improving margins without adding headcount.
The move reflects Basware's effort to apply AI to service delivery rather than limit it to task automation. It began with billing processes, then expanded into resource management, implementation work and project documentation.
Mark Johnston, Chief Customer Officer at Basware, said the effort began as an operational problem inside the services organisation. “Our services organisation was low-margin, slow to scope, slow to staff, and I couldn't tell you whether the software we were running the business on was actually working for us,” he said.
Basware provides invoice lifecycle management software to thousands of customers worldwide. In a case study released with Kantata, it said it wanted a clearer view of how AI could improve operational performance and customer outcomes.
Johnston said the company concluded that the bigger opportunity was in extending the judgment and experience of its staff across more engagements. “AI can do the heavy lifting on routine work, but the real opportunity is using it to scale the expertise of our people,” he said. “Every customer environment at Basware is unique, and that's where our people make the difference. What if we could take what our best people know - their judgment, their patterns, their experience - and make that available consistently across every project and every customer? That's where AI gets really interesting, and the direction we are moving.”
The first area of focus was billing automation. Basware then expanded AI use within the Kantata system to help services managers staff projects as timelines and requirements changed, aiming to match staff to accounts while protecting delivery quality and margins.
It also used AI to handle repetitive elements of software implementation work and generate project documentation for customers. Basware said that shift allows professional services teams to spend more time on customer-facing work rather than administration.
Capacity gains
Basware gave examples of how the changes have altered working patterns. Johnston said customer success managers who had been spending four to five hours a day researching before customer calls are now using that time in direct conversations with customers.
It said that extra capacity could allow it to double revenue from its existing customer success management team without increasing headcount. In support operations, an AI agent is now handling roughly half of the 10,000 monthly tickets submitted by non-paying customers of Basware's customers.
Basware said those tickets are being resolved more quickly and with clearer responses, contributing to higher satisfaction levels. It did not disclose financial figures for the investment or margin improvement.
The announcement comes amid broader scrutiny of how companies measure returns on AI spending. Research from EY, cited in materials accompanying the case study, found that while half of UK chief executives say AI is lifting productivity, only 14% report having comprehensive, real-time visibility of AI costs and return on investment.
Kantata said the Basware deployment shows how services businesses can move from isolated experimentation to day-to-day operational use. Michael Speranza, Chief Executive Officer at Kantata, said pressure on professional services teams has pushed the debate beyond simple automation.
“In professional services, automation is table stakes. Expertise is the differentiator. And the challenge is scaling it without diluting what makes it valuable,” he said. “Basware is showing what that can look like in practice, using AI to take the work off its experts' plates so they can spend more time doing what customers actually pay for - solving problems, applying judgment, and delivering outcomes.”
Basware's account adds to a growing number of examples of companies trying to tie AI projects to operational metrics such as utilisation, staffing speed, support volumes and service margins, rather than broader claims about efficiency. In this case, its argument is that the main gains come from reallocating employee time to areas where customer-specific expertise matters most.
That distinction is likely to resonate with services businesses that depend on specialist staff and are under pressure to increase output without expanding teams. Basware said its experience suggests the economics of AI in services depend not only on automating routine tasks, but also on deciding where human judgment has the most value.