Earlier this year at the DPW Amsterdam event, Sean Galea-Pace of CPOstrategy sat down with Danny Thompson, Chief Product Officer at apexanalytix, to discuss the art of developing trust amid significant innovation in procurement. We’ve republished Sean’s write-up here.

Trust.

apexanalytix needs to build quite a bit of it. As a company which protects $9 trillion in spend and prevents or recovers more than $9 billion in overpayments annually, its client portals actively support over eight and a half million suppliers.

Indeed, apex has revolutionised recovery audit with advanced analytics and the introduction of first strike overpayment and fraud prevention software. Today, apex is a leading global force in supplier management innovation with apexportal and smartvm, now the most widely used supplier onboarding, compliant master data management, and comprehensive third-party risk management solution in global procure to pay. With over 250 clients in the Fortune 1000 and Global 2000, apex is dedicated to providing companies and their suppliers with the ultimate supplier management experience. A big part of that experience is based on building trusted supplier-buyer relationships.

Danny Thompson is the Chief Product Officer at apexanalytix and has been with the organisation since July 2015. Now in his third role with the company in eight years, Thompson reflects on his journey with the organisation with positivity. “I came in as a product manager working on our portal product,” he tells us. “And after a short time, because I was a former customer, at Pfizer and International Paper Company, and was an internal voice of the customer, they ended up having me drive messaging with marketing. Recently, we hired a great new leader of marketing who has taken that over fully so I’m dedicated full time to product again. So it’s been a great experience for me.”

Gen AI surge

One of the hottest topics on the CPO agenda in recent months has been ChatGPT. Wherever you go within the industry, you’ll likely find a conversation being had about the technology’s possibilities, as well as perhaps its limitations or challenges – and Thompson is equally keen to explore.

Danny Thompson speaks with CPOstrategy at DPW Amsterdam 2023
“There is certainly a lot of attention being paid to gen AI in the industry, and within our company as well,” says Thompson. “I think it’s because of the shock value of ChatGPT hitting the world and people are really stunned by its ability to interpret natural language and come back with really good information in response to questions that are being lobbed at it. There’s a lot of excitement around what it could do as well as what other generative AI solutions can do to help solve procurement, supplier risk and supplier information problems. We are making progress, and have introduced some generative AI functions, but Generative AI presents some challenges right off the bat that we are working hard to solve as quickly as we can.”

One of these issues is the hallucination problem that is being questioned within the space. This is where AI tools like ChatGPT lack factual support for some of the information provided. “There’s a statement at the bottom of the page which states you can’t rely on results being factual,” Thompson affirms. “When it comes to supplier information and risk management, that’s a problem.”

Managing Risk

And it is such an important sticking point that Thompson stresses when it comes to supplier risk information, it is about being careful that the usage of generative AI, in its current state, is used for guidance rather than fact-finding. “Another challenge is around leakage of sensitive information combined with contamination of sensitive or important information,” reveals Thompson. “We have a database of golden records for 90 million suppliers who are doing business with Fortune 1000 and Global 2000 companies. That is the best information we’ve been able to accumulate about suppliers and their relationships as a supplier to large companies. Some of that data is publicly available and some of it is more sensitive. We want to make sure we’re not loading that sensitive information into a generative AI function that might allow random people to access that information. We’ve got to be careful about that leakage of data.”

The opposite is true, as well. Thompson reveals that his team asked the generative AI-tailored questions which they assumed would be pulled from their own database. The findings were less than ideal. “The responses had been contaminated with public information which was full of inaccurate data,” he tells us. “We’re figuring out how to draw those boundaries, as well—to protect sensitive data while also preventing contamination.”

Trust first

This showcases the importance of trust once again to an organisation like apex. The companies it serves are moving significant sums of money around and the potential risks are sizeable. For Thompson, there can be no greater responsibility when using AI tools. “The data must be either highly accurate or at least they understand the degree to which it’s not,” he says. “If you don’t understand that level of trust you can have in it, then you shouldn’t be using it yet.”

With an unprecedented amount of technological innovation at procurement’s fingertips, the industry is evolving at a rapid pace. It’s placed at a unique moment with digital transformation being swept up throughout the space. While this brings obvious advantages such as time and cost savings, it also means increased cybersecurity threats. “There are more threats coming in as a result of AI,” says Thompson.

“One of the biggest challenges our clients us our solutions to solve for is fraudsters trying to take over a supplier’s account and intercept their payments by submitting fraudulent account change requests. One of the typical ways companies catch these is very often the request is coming through very poorly formatted emails with bad grammar. But what we’re seeing is the bad guys have started using generative AI to create really convincing bank account change requests so there are increased threats to be aware of. But this increase in the availability of information is also make easier the whole process of knowing your supplier and knowing the risks associated with them. And Generative AI is going to allow you to quickly get help to understand how to mitigate a given risk much faster and easier than it’s ever been before.”

Interested in learning how apexanalytix leverages the latest advancements to help your company manage risk? Contact us now to start the discussion.

About Danny

Danny Thompson is the Chief Product Officer at apexanalytix. He collaborates with P2P clients to understand their needs and challenges, and to ensure apexanalytix solutions drive better outcomes to meet those requirements.

On a 25-year foundation of experience in P2P and shared services experience, process optimization leadership, and ERP system expertise, Danny collaborates with industry analysts and co-innovates with apexanalytix clients. He is responsible for the definition and communication of innovative services and software that address risk, improve P2P controls, and anticipate the next generation of process automation and efficiency.

He previously was VP of Product Management at a global business-to-business e-invoicing firm, and global process driver for Invoice-to-Pay at Pfizer.