Rethinking existing monitoring systems to exploit potential

Every bank deals with financial transactions. The number they deal with varies greatly depending on the orientation and focus of the bank. Controlling is one of the major tasks in these financial transactions. It must be performed quickly and to a very high standard. It is the only way to ensure compliance with all legal requirements.

Every year, money laundering and fraud cost banks billions. In addition, institutions such as the EU or the German Federal Financial Supervisory Authority (BaFin) impose severe penalties if these controls are not carried out in accordance with procedures. There are also other areas where action needs to be taken alongside the actions and specification in regard to the area of know your customer, for example.

One example of this is the embargo analysis of cash transactions. This involves scrutinising every financial transaction on the basis of embargo and sanctions lists. The process involves assessing whether the financial transactions are directly related to providing financing and/or assistance when transferring specific goods to countries, persons or entities.

Banks already have systems in place to do this. These systems check the transactions based on rules and specifications, enabling them to quickly ascertain which ones can be approved without issue and which ones have to be checked manually.

Over 6.4 billion transactions that need to be checked make automation a must

Complex specifications mean that some financial transactions still need to be checked manually – at the expense of great time and effort. In typical cases, manual control is needed in about two per cent of all transactions. Once you factor in a (steadily increasing) volume of 6.45 billion transactions per year , this works out at 129 million transactions that need to be checked manually each year. Seeing as each transaction takes 30 seconds to check on average, this means that over 200 years are spent checking them.

This challenge can be overcome and the number of financial transactions that need to be checked manually can be significantly reduced by using modern artificial intelligence methods such as machine learning algorithms. These processes can be used to identify entities – such as sender, recipient or intended use – which in turn can be checked against the regulations (the embargo and sanctions lists).

adesso’s approach ensures success in next to no time at all

adesso provides a platform to industrialise this process and to build on established ones used by our customers. This platform provides microservices that can be adapted to each payment format and the individual requirements of the banks, meaning it can also be continuously developed (trained).

A separate neural network is built for each payment format so that the specifics of each one can be accounted for. To put it in simple terms, a neural network is a connection of different nodes to process data and thus ensure a high recognition rate for the individual entities. The machine learning models that are customised according to the customer’s individual requirements are the customer’s property and remain the respective bank’s intellectual knowledge.

The quality and speed of processing that this approach achieves are superior to those that are based purely on rules. Moreover, this approach can also be used in highly regulated environments. Other implementation and project approaches are chosen depending on the individual requirements and circumstances.

In general, our approach is based on an agile approach. The data, AI models, rules and specifications are tested in a minimum viable product (MVP), which is developed within three months, before being implemented in a target concept. The MVP is then put into production. The focus here is on transparency as well as high quality and safety, and every action is discussed and coordinated with the customer.

Conclusion

In summary, we can say that approaches based on machine learning algorithms achieve a significantly higher level of quality, reduce the amount of manual effort needed and quickly achieve a return on investment as a result.

Would you like to learn more about exciting topics from the world of adesso? Then check out our latest blog posts.

You can find more blog posts on AI here

Picture Tim Strohschneider

Author Tim Strohschneider

"Complicated contexts presented simply" - Tim's motto as a problem solver in the Line of Business Banking motivates him every day, with the customer in focus, to make all topics related to intelligent automation and artificial intelligence accessible in the business context. Over 15 years in the IT world, consultant at heart, leader and innovation driver - he describes his view of current developments here in the blog and on his LinkedIn page.

Save this page. Remove this page.