Commerzbank is piloting a fintech solution that will see it automate 80% of its “first line of defence” compliance checks of its trade finance processes by 2020.

The bank has announced it is working with Conpend, a fintech firm that uses optical character recognition (OCR) and machine learning to digitise physical trade documents, recognise patterns and flag deviations. The firm also conducts automatic document checking and sanctions screening.

Speaking to GTR, Marc Smith, Conpend founder and MD, says the two have been working together since May to integrate the system into Commerzbank’s back office. The bank, he says, is now live with the software in production in Hamburg, where it is carrying out the pilot.

At first, the pilot will see Conpend automate Commerzbank’s anti-money laundering processes, and, after an evaluation process, the bank plans to also implement further checks in trade finance next year.

Commerzbank says in a statement that the planned 80% automation refers to business operations only – the so called “first line of defence” – and does not affect any downstream internal oversight, compliance and audit processes.

It is expected that the tool will radically change the way hundreds of the bank’s staff work today, Smith says. Describing the solution, he explains: “Today, in a manual situation, employees get documents and then they have quite an extensive checklist that they have to go through manually, and record that they actually do a check. So they might have an Excel sheet in which they enter when they did a check and attach a screenshot of the result or any other evidence that this happened. With the new solution, all of that will now be automated, so they can upload documents and then get a report. They can then see, here it was flagged green, and here was an issue, and in the tool itself, they can then handle that issue. It will obviously save them a lot of time.”

He explains that the pilot stage is more about Commerzbank reviewing its internal work processes than piloting the technology itself. This includes the bank updating its procedure manual, while also documenting the training process so that it can be scaled up to the broader organisation.

According to Smith, Commerzbank will potentially start to see the first time saving as soon as this year, after the pilot has ended.

Enno-Burghard Weitzel, head of product management trade services at Commerzbank, says the decision to adopt Conpend’s solution came as the processing of trade finance transactions was becoming more complex and prone to higher risks, and the bank was struggling to keep pace with the increasing regulatory and market trends using its manual systems.

Commenting on the partnership, he says: “Our aim is to focus the expertise of our trade finance specialists to the crucial and complex parts of the business, while using artificial intelligence to improve efficiency and further optimise risk controls. This is not only a long-term plan, but something we are implementing now to enhance client experience by significantly reducing the time for transaction processing and the associated costs.”

ING is so far the only other bank that has worked with Conpend to implement parts of its solution. According to Smith, the fintech firm is currently running proofs of concepts with other banks in Asia and America. One of the projects was announced at the Sibos conference this week and involves working with Scotiabank and CGI on a proof of concept using robotic automation to capture, identify and classify documents as part of the CGI Trade360 transaction flow.