Artificial intelligence in food production will likely conjure up images of robots in factories packaging products or autonomous quality assurance devices on the assembly line. But there’s another area where AI is having a huge impact on efficiency and profits: the back office.
Accounts payable (AP) has become one of the top automation priorities due to its document-heavy processes and high consumption of time and resources, which often divert employees’ attention from more customer-facing and value-generating tasks. In fact, recent research shows that the AP automation market is expected to grow from approximately $6 million in 2024 to $17 million by 2032, highlighting its importance in the digital transformation journey.
Mars is one such company that has made the decision to prioritize AP automation to keep up with its continued growth. Maxime Vermeir, Senior Director of AI Strategy at intelligent automation company ABBYY, helped Mars standardize its AP processes across its global offices. Vermeir has a decade of experience in product and technology, and his expertise in AI helps drive business solutions and transformation initiatives.
FOOD ENGINEERING spoke with Vermeir to get a first-hand look at the challenges, strategies, and results of implementing AI in accounts payable.
FOOD ENGINEERING: What were Mars’ most pressing pain points before automating accounts payable with AI?
Maxime Vermeir: Their accounts payable department encountered challenges common to any organization without accounts payable automation: extensive manual data entry resulted in errors and inconsistencies, ultimately requiring a significant amount of extra work that could have been avoided.
These are obvious targets for AI-powered improvement, but their rapid growth and global scale presented another challenge. Mars would have needed to hire 50-75% more people to manage its invoices and needed to find a way to meet that need through automation. Additionally, its existing accounts payable staff is spread across many different countries, each with its own tribal knowledge of best practices and regional differences.
In short, their staff got bogged down in data entry instead of steering the spaceship to boldly go where no one had gone before.
FE: How did AI meet this need for standardization? What was Mars’ strategy?
MV: It was important for Mars to keep the business side of its organization in the loop throughout this process. This was the catalyst for the success of its strategy, which ultimately involved creating a baseline standard that summarized how its accounts payable processes should look at a high level, taking into account regional variations. Mars created two documents, each over two hundred pages long, outlining the details and nuances of its accounts payable functions across regions.
Choosing the right AI solution was like designing a new Iron Man suit: combining cutting-edge technology components into a robust system that could solve this complex problem. They chose a low-code, cloud-based Intelligent Document Processing (IDP) platform that leveraged natural language processing (NLP) and machine learning, which allowed them to aggregate invoices from over two thousand different vendors into their ERP system. With NLP enabling semantic analysis to contextualize AP language while machine learning enabled AI models to be trained on an infinite range of document formats, Mars was able to extract valuable data consistently with speed and accuracy.
Using this IDP approach, Mars has socialized AI-enhanced billing-payment processes across 20 global markets in 14 different languages.
By integrating AI into accounts payable, Mars was able to pursue strategy and value with staff who would otherwise be assigned more monotonous back-office responsibilities.
FE: What were the benefits of this initiative?
MV: By integrating AI into accounts payable, Mars was able to implement strategy and add value with employees who would otherwise have been assigned to more monotonous administrative tasks. IDP significantly accelerated invoice processing and achieved higher straight-through processing (STP) rates, meaning that a large portion of their documents were able to be processed without any manual intervention from human employees.
Reducing this heavy document burden has freed them up to focus on judgment-based goals, such as disputed transactions and other value-added activities that cannot be performed independently. Beyond the obvious efficiency and revenue benefits, this has also meant a reduction in monotonous and tedious tasks for employees. With data from a recent survey revealing that 92% of employees spend up to eight hours a week combing through documents for information, that’s no small amount; it could mean the difference between employees enjoying their jobs and completely burning out.
FE: Should Mars have done anything differently?
MV: The Mars implementation was both an anomaly and a master lesson. We saw an abnormally fast pace of growth, and so they had to be particularly strategic in implementing their automation strategy.
They’ve taken a big step forward in ensuring that their finance department is involved throughout the implementation process. This is critical to ensuring the long-term effectiveness of AI. It can’t just be about technology. It has to translate into business value and solve a real challenge.
While this answer may seem superficial, the only thing that really comes to mind is to start earlier. Getting behind the wheel of the DeLorean and accelerating in a strategic, data-driven direction gives you more time to interact and adapt to the technology, which ultimately gives you a head start on the future of operational excellence. With AI and automation, you can only really know if it works once you’ve had time to interact with it and see how it fits into your entire organization. Without that hands-on experience, it’s hard to build a solid foundation.
I would advise organizations seriously considering automating AP processes to be careful in their choice of solutions and implementation partners, and to make data-driven decisions about where AI can play the most significant role.
For example, intense efforts to organize complex and diverse processes into formalized documents could benefit enormously from data-driven tools like process intelligence. By collecting data at every step of a workflow, process intelligence provides the most comprehensive visibility into how processes are executed end-to-end. This enables efficient and accurate representations of core workflows, which could significantly accelerate initiatives like Mars’ 200-page AP workflow documents.
When I recently joined Mars at a digital summit on AP automation hosted by SSON, 70% of attendees said they were evaluating or learning how to use AI in AP, and I think that strongly suggests that this growing trend is not going to slow down anytime soon. Missing the AI train could be a recipe for disaster for any food manufacturer.