A new fintech company, Tradesun, has entered the trade finance space, launching an AI-driven solution to digitise and automate the processing of trade documents. It is currently working with nine banks across the globe to pilot its cloud-based application.

The company’s AI solution combines optical character recognition (OCR) and natural language processing (NLP) technology to read and extract information from a wide variety, type and quality of complex trade documents. It simultaneously scans for suspicious activity and flags risks of regulatory and fraudulent infractions by checking all data against compliance and business rules.

Tradesun was launched as a company last year by the founder of Data Skill, a technology firm that has been running since 1981, initially providing a contract system for the US department of defence.

In 2011, Data Skill started developing an AI solution for multiple industries. Since 2017, it has been working to implement the system with one of the largest trade finance banks in the world, having already run pilots with the bank over the previous year.

Now incorporated officially as Tradesun, the fintech firm is fully focused on trade finance. Its new solution differs from the previous version in that it is hosted on Amazon Web Services and made available to clients under a software as a service (SaaS) model.

And while the previous solution was powered by IBM Watson, Tradesun has now developed its own algorithm that is focused specifically on trade documents such as letters of credit, bills of lading and insurance forms.

“We’re now 100% focused on trade finance, and rather than having customised solutions for individual companies to give them a competitive advantage, we are making this available to the whole trade finance banking industry,” Tradesun’s CEO and founder Nigel Hook, tells GTR.

“Over the last few years, working with one of the world’s largest trade finance banks, the solution has been trained on loads of trade documents, and because it’s an AI system, it gets smarter over time. That’s why it’s very accurate in the classification of types of documents and the extraction of the trade data and matching between documents. It’s now reached a point where we feel very comfortable that it’s getting excellent performance right out of the box for new bank customers.”

Within many banks, the document checking process is still a manual, labour-intensive job. According to Hook, the Tradesun solution has so far amplified the users’ throughput by three to seven times.

“Using AI automation, the number of documents a bank can check every day is multiplied many times,” he says. “We make the checkers into ‘super checkers’. What is also important when you talk about the interception of fraud activities is the thoroughness of being able to extract all the data and check for red flags and sanctions breeches in real time. Finally, we are now helping the banks’ customers, because the turnaround is faster.”

Production later this year

Tradesun is currently running pilots with nine banks across six continents and has also been working on a trial with the Monetary Authority of Singapore to help flag suspicious transactions. The company anticipates being in production with the first clients later this year.

Tradesun is not the only young fintech player using emerging technology to help banks automate what is today an extremely paper-based and manual industry, with similar solutions having been presented to the trade finance market over the past couple of years.

Traydsteam, for one, launched its AI solution in 2017 and went live with its first bank client last year. It is now working with Finnish technology firm Nokia and three banks – OP Group, SEB and Standard Chartered – to pilot the application, which the company says can reduce the average time it takes a bank to conduct a document check from four hours to minutes.

Conpend is another company operating in this space. It recently announced it is working with Germany’s Commerzbank to integrate the system into its back office, with the bank claiming to be able to automate 80% of its “first line of defence” compliance checks of its trade finance processes by 2020.

Meanwhile, Hook at Tradesun sees it only as a positive for the wider trade finance market that a growing number of players are now helping banks implement AI for the processing of trade documents.

“It’s going to help trade finance in underserved markets in the world, which are highly risky, and so, they have trouble getting trade finance,” he says. “But with an application like Tradesun, it simplifies everything for importers and exporters and banks on both sides. So it’s going to generate a lot more financing in countries which have struggled in the past.”