U.K.-based digital banking startup Revolut is expanding at a rapid clip.

The company boasts more than 8 million customers across multiple continents. It operates in Europe and Australia, and last fall launched publicly in Singapore and in beta for U.S. customers.

Revolut’s efforts to expand means meeting regulatory requirements in each country in which the company operates. To improve efficiency of compliance workflows, Revolut has resorted to automation to help solve this problem by using a tool developed by London-based regtech firm ClauseMatch, which also works with large financial firms like Barclays and Italy’s Intesa Sanpaolo.

For the past 18 months, Revolut has been working with ClauseMatch to map regulations to policies and provide evidence of compliance in the jurisdictions where Revolut operates, a task that would take thousands of hours to carry out manually, according to the company.

“There’s a couple of aspects to it: the ability to coordinate approval and review from the same location, and the ability to organize and connect our policy to the rest of our risk and compliance data points,” said Nicholas Melas, senior global policy manager at Revolut.

According to Evgeny Likhoded, CEO and founder of ClauseMatch, the platform allows for real-time content collaboration and workflow management, and lets users map the content across the platform. The tool also uses natural language processing and machine learning to suggest relevant content.

See also: With Mastercard, Revolut aims for US Market

ClauseMatch allows Revolut to coordinate input, approvals and workflow in the same location without having to provide users with links to different passwords; it also lets the company automate the policy-approval process. This offers two benefits: consistency of approach and the ability to make changes to policies with minimal legwork.

While ClauseMatch streamlines policy approvals and changes, it doesn’t eliminate humans completely, emphasized Melas. Instead, it takes away menial tasks, allowing staff members to focus on more complex roles, including oversight and verification.

“It’s a mix [of humans and machines],” Melas said. “For example, we don’t want people to be spending their time copying paragraphs of material; the automation is extracting approved policy from the ClauseMatch platform into our back-office systems for association.”

Looking ahead, Melas said he foresees other use cases for the technology to drive workflow efficiencies.

“There are a lot of ‘business as usual’ activities that seem simple, like reviewing a policy or reviewing a document for compliance and policy, but it’s done extremely poorly in large swaths of the industry because it’s manual, and there is no quantitative assessment of whether something is done well,” he said.

Banking Automation Summit, which takes place from June 1-2 in Miami, is a unique opportunity to share insights, trends, strategies and best practices on back-office automation in financial services with the industry’s leading practitioners.