![Case Study: Quick fraud detection](https://www.widas.de/wp-content/uploads/2022/12/AdobeStock_295766804.jpg)
The task
Nowadays, a customer expects their bank to know and understand their individual current situation and to refer to it accordingly in their communication with them. In addition to the current product portfolio, the customer approach should therefore be tailored at least to the available balances, the customer’s current transactions and existing customer data.
The requirements of our client – a large bank – were even more complex:
To approach the customer optimally from a service, marketing and sales perspective, information and data from all channels were required in real time, i.e. in real time. The entire customer journey was to become transparent. By linking these data streams with information already existing in the bank, an optimal profile of the individual customers can be read out. For example, with the aim of identifying possible migration to competitors in good time and counteracting this with appropriate communication.
Overall, the bank had the satisfaction of its customers in mind and therefore also wanted to expand the services that could be called up. For example, by updating the account balance immediately after the booking, which can also be read on the smartphone without time delay.
The added value
The smart banking platform proved to be a quantum leap for the bank’s marketing in particular: the timeliness and quality of the customer approach is incomparably better and above all – uniform across all channels offline, online and mobile. Numerous service topics now replace the content purely around product sales.
The bottom line: happier and more satisfied customers of the bank. And a new revenue situation, which our client is also happy about.
![Case Study: Quick fraud detection](https://www.widas.de/wp-content/uploads/2022/12/AdobeStock_516805712_02-2.jpg)
![Case Study: Quick fraud detection](https://www.widas.de/wp-content/uploads/2022/12/bigdata-service_widas.jpg)
The solution
With an individual Big Data solution beyond extensive pre-conceived models, we were able to store incoming data without a structure, to quickly retrieve information from the database and to answer many queries in parallel, as well as to identify similarities between customers and types of use within a short time.
The technologies used
Hadoop, Cassandra, Storm, Kafka, ElasticSearch