Banks facilitating trade finance have to deal with the realities of paper-based documentary trade finance. This is an area ripe for innovation, and certainly one where artificial intelligence (AI) in particular can play a significant role. Conpend founder Marc Smith and CEO Torben Sauer explain how their partnership with ING Bank has changed the game with respect to trade efficiencies.


ING Bank, one of the world’s most tech-savvy banks, is constantly looking to see where innovation can add efficiencies: not least in trade finance. One area that has resisted innovation – and especially digitalisation – is the checking of documents such as bills of lading and letters of credit (LCs). Supply chains – especially those involving small emerging market manufacturers in the poorer countries – not only still rely of paper-based documentation but will do so for the foreseeable future.

Banks on their own cannot solve this problem. But they can apply fresh thinking, which is what ING has done. The bank accepts that electronic documentation is simply not viable for certain supply chains, so instead wants to make the existing process better. Working in partnership with trade software specialists Conpend, the bank has deployed intelligence (AI) to address key aspects of documentary processing. These include the checking for compliance against International Chamber of Commerce (ICC) rules, as well as anti-money laundering (AML) guidelines and regulatory sanctions screening, including for the Office of Foreign Assets Control (OFAC) in the US.


Transforming documentary checking

ING implemented Conpend’s TRADE AI app, which automatically checks documents against inputted rules and regulations. As has been the case with trade finance departments for decades, documents arrive by envelope and are then scanned and sent through a secure file-share for processing to their Shared Service centres. Yet, while operatives had to once laboriously read all the documents before compiling a compliance report, AI now scans the document for key words and phrases against inputted rules and stipulations. If an anomaly is spotted, a query is raised and one of the operatives checks the document.

Such technology has the power to convert all documents – copies or originals, clear or unclear papers – into machine readable words through optical character recognition (OCR). The app is also self-learning: applying learning from previous documents to all future documents. The result is a process that is faster, more accurate and more precise than anything the human eye can replicate, allowing it to produce analysis as well as recommend actions and solutions. And all while constantly improving.

Meanwhile, operatives are freed from the drudgery of document reading in order to spend more time checking the queries identified by the AI solution – adding further to the efficiency gains.

For ING the objectives are clear: faster turnaround times, reduced errors and improved interaction with clients for day-to-day activities, all of which the TRADE AI app achieved. An important further benefit has been the “employee experience” – including increased job satisfaction at the processing centres, not least because the document checking process was repetitive, boring and time-consuming.  Also, manual processes are by definition error-prone, making the AI process additionally beneficial from the point of view of risk reduction.

“The partnership has been a journey, and it has been approached as one right from the start, with no expectations that TRADE AI would be installed and immediately save time,” says Dermot Canavan, Trade Operations Manager at ING Bank. “That said, the app has delivered more than we expected quicker than we expected.”


AI and due diligence

And AI capabilities are not restricted to trade documentary checking: multiple bank processes can benefit, including areas where due diligence is necessary but remains a highly detailed yet mundane documentary-checking process. For instance, with punitive fines being handed to banks for even minor KYC or AML transgressions, it is possible for AI to simplify the time-consuming process of onboarding customers and suppliers. Again, the focus is on automating tedious tasks, minimising human error and allowing banks’ compliance professionals to focus on risk management.

Certainly, the partnership between Conpend and ING Bank – forged through the development of trade documentary checking – is now entrenched, opening up other bank processes that could benefit from AI.