Intelligently predict risks: with AI to the next level of factoring services

In receivables financing, it is crucial for the factor to react quickly while taking decisions that will have an immediate impact on its profitability or even upon its business continuity:

1. How good is the “quality” of receivables a client would like to have financed?
2. Is the agreed financing limit too low or have I exceeded on risk appetite?
3. Is what I new about my customer a year ago, a month ago or a day ago still apply today, now or perhaps tomorrow?

Such far-reaching decisions have to be made repeatedly, from receivable to receivable, from customer to customer, but in contrast to lending, they have to made in minutes or even in seconds. Up to now, external service providers, such as credit insurers or rating agencies help in providing information about the credit worthiness of clients and debtors. There have also been initial attempts to create solutions using A.I. or blockchain to help factoring organizations taking the right decisions. But in an increasingly interconnected supply chain, i.e., business world, organizations are dealing with a complex set of relationships. It thus becomes increasingly important to also understand which customer is dealing with which debtors, which debtor is dealing with other debtors or which customer is dealing with other customers, the total factor’s eco-system.

Big picture

Understanding the eco-system means, identifying opportunities but also possible risks, how to determine in advance whether one of the players is exposed to certain risks or even fraud. This is where factoring solutions must aim for the next level of security and thus also trust, this is when we aim for Artificial Intelligence (A.I.) and Machine Learning (M.L.) as supporting technologies equipped with the performance and overview that goes far beyond the „simple“ observation of invoices: The big picture is once again in demand, where self-learning systems provide a holistic overview of the current, critical states of all parties involved – and in real time. This is the only way to decisively push the factoring industry, which is undoubtedly in a significant upswing.


In essence, it is about identifying debtors risk in real-time! If information from credit insurers is used for this purpose, it has traditionally been based on historical data and in the case of fraud, for example, one can only react. In case of doubt, it is perhaps already too late and a whole cascade of processes follows in order to prevent something even worse. It would certainly be ideal to know in advance whether and possibly also when a customer is in danger of dropping out of the normal daily business. An AI/ML – based application can detect past and current payment behaviour of customer’s-customers (debtors) and make predictions for the future when using real-time data. Other factors should also be taken into account in order to get as comprehensive a picture as possible. These include risky sectors and, of course, the data from credit insurers already mentioned. The system would also act on the basis of known or newly learned fraud patterns. Ultimately, such checks would be automated and users would always be shown the current state of affairs in the form of a dashboard. And as I said, all of this would be network-based and take into account all the relationships between the actors known to the system.

Multi-relationship network

The graphic below shows how such a multi-relationship network can be displayed in real time. Using the colours (red, green, yellow) as well as the symbols (thumbs up/down), the current status of the respective debtors can be observed. This is not only about the (early) recognition of risks and dangers, but also about possible opportunities. Where can potential be positively exploited? With which customers is it worthwhile to further intensify business relations? As already mentioned, the updating of statuses is absolutely dynamic and is generated with the help of AI on the basis of various data. In short, making use of such data can be invaluable for your own factoring business.

Of course, anyone working with such an AI-based tool will inevitably benefit from efficiency advantages. You will be able to make decisions with (even) fewer resources and coordinate processes (even) better. As a result, factoring providers would have lower costs for their services. This in turn would have a positive effect on the barriers to entry for new market participants. However, increased efficiency also means increased speed with better quality decisions. In addition, complexity is reduced overall. All these factors would also have advantages for the end customers: Besides the fact that they would also benefit from faster decisions, it is conceivable that they would also be offered individual conditions based on their respective historical performance – as a kind of reward.


Conclusion: The factoring industry is facing a real paradigm shift with a far more comprehensive use of AI than before. We are talking about intelligent pre-warning systems that will be able to automatically recognise a multitude of risks and report them in a timely preventive manner, even in the complex relationship network of the end customers. What’s more, such a system will reveal not only the risks but also possible opportunities and potentials that would hardly be recognised with the help of classic tools and/or only with a corresponding delay.


So it is time to offer a feature called multi-history & opportunities in addition to the usual services such as multi-currency and multi-language!


There are also other ideas on how AI can be used to improve factoring services. These include the following*:

      • email support in composing and responding to emails to prospects and customers.
      • Monitoring social media for customer interactions and feedback, complaints and enquiries, and analysing sentiment accordingly.
      • Personalised customer support based on each customer’s preferences as well as history
      • Automated ticket management by categorising and prioritising customer support requests with suggested solutions based on historical data and common issues
      • Generation of knowledge bases based on past customer interactions
      • Simulation of realistic customer support scenarios for internal and external training purposes

*Mark Mandula, 2023: Some Other Ideas on how to use Generative AI in your Factoring Firm