Related News

Traydstream hires CTO to boost innovation

Xalts founders on trade finance digitalisation after Contour acquisition

New template may spur open account trade finance distribution

Surecomp integrates Cleareye onto trade platform

Swift says CBDC testing for trade payments showing positive results

HSBC and IBM have developed a cognitive solution to automate and digitise trade finance documentation.

The solution, which is already in use in Hong Kong and the UAE, uses IBM robotics technology to analyse documents, digitising and extracting the relevant data before feeding it into HSBC’s transaction processing systems.

The aim is to remove the labour intensity from trade finance. HSBC’s global trade and receivables financing (GTRF) team processes more than US$500bn in documentary trade each year, meaning more than 100 million pages must be manually reviewed and processed.

Using artificial intelligence (AI) to do this frees up the bank’s staff, and also removes human error from the equation. The bank is currently using it to analyse English-language import and export bills. It can read documents in Chinese, French and Spanish.

HSBC’s head of client and transaction services, GTRF, Jonathan Moore, explains to GTR how the solution works: “The solution is a two-step process. Once all the documents have been scanned – on average 40 pages per transaction – the software reviews each document, categorises and files each page, the application form, bill of lading, packing list, etc, are labelled accordingly.”

He continues: “Once the documents have been sorted and indexed, the software reads each page to extract the required data based on defined rules. The key data is automatically entered into our system – typically 65 fields need to be filled for each transaction. Aside from the volume of documents we handle and the level of complexity, the challenges with automating the process is the unstructured nature of the documents and the underlying quality.”

Roger Welch, an IBM architect who helped build the solution, says that “no trade finance solution has come as far or done as much as this”.

He adds: “The problem is how to capture semi-structured documents with highly variant content through an analogue process, and no-one has the perfect answer.”

Automation has been a hot topic in the paper-intensive world of trade finance in recent years. Banks have been investing heavily in AI solutions and developments, with fintech companies also developing solutions for the market.

Earlier this year, in one notable example, fintech company Traydstream was launched, with a solution that uses AI to digitise and automate trade documents, and automate regulatory compliance screening.

Other companies connected with trade finance are using AI to combat financial crime. In May, payments and compliance systems provider Pelican launched a trade-based money laundering (TBML) detection solution which uses AI technology.