case study

Case Study: Online Bank

Find out how a renowned German neobank can not only optimize its processes by digitizing its letter mail and automatically recognizing the corresponding content, but also meets high compliance requirements.

‍

challenge

Regardless of their numerous digital offerings, banks still receive many letters every day. This includes letters from customers such as account cancellations, but also documents, correspondence in insolvency proceedings or seizure notices, all of which must be processed in accordance with compliance and to meet any deadlines. No letter may be left “by mistake.” To meet this challenge, the online bank has digitized the letters itself. The employees then viewed the digital documents and added them manually to the respective software. This procedure involved an enormous amount of time and therefore high costs, because after all, the bank receives more than 10,000 letters a month — and the trend is rising. On the one hand, to ensure that every letter is automatically forwarded correctly and, on the other hand, to reduce the associated enormous effort, the bank was looking for a professional service provider for the entire processing of incoming mail.

‍

requisitions

The service provider should pick up the mail from a bank location or directly from Deutsche Post, digitize it, read out data and then correctly assign it to the respective work processes based on the content of the letter. For this, the service provider received more than 100 relevant terms, from which the corresponding process is derived. In addition, the service provider should filter out letters from around 40 special senders and make them available to the bank unopened. In addition, depending on their content, certain letters should also be destroyed after digitization, others should be physically archived and others should be handed over to the bank in their original form.

‍

solution

To ensure that the recognition of work processes based on the respective letter content is largely automated, Caya uses machine learning technologies that were initially trained manually. Specifically, the letters that contained one of the relevant terms were marked and linked to the respective processing logic. Depending on the keywords they contain, the documents are transferred to various folders on the Caya platform or via interfaces to industry solutions. Caya saved these rules for all terms and was therefore able to automate editing almost completely. Since January 2022, Caya has picked up letter mail several times a day with her own courier at a bank location or from Deutsche Post's mail distribution center and brings it to one of its scanning centers. Here, letters are first sorted out which the bank would like to receive unopened, such as letters from the works council that return with the next courier trip. The remaining shipments are digitized, read and typed using trained machine learning technologies in order to feed them to the appropriate processes. Finally, depending on the content of Caya, the letters are either destroyed in accordance with data protection regulations, physically stored in a long-term archive or also returned to the bank.

‍

“As a service provider, Caya processes letters, including automated opening, preparation and scanning.”

‍

upshot

Caya processes the bank's letter mail within a few hours and transfers the digital documents, including the extracted data, to the associated processes during this time. In this way, the customer was able to significantly relieve his employees. In addition, the new process ensures that any deadlines are met and that all letters are processed in accordance with compliance. This is also supported by detailed reporting and dashboards, which Caya provides to its customer. These provide information at any time about the number of incoming letters, where they are and what processing status they currently have. In this way, the bank is always able to provide information and can intervene accordingly if a process stalls.

‍