Colossal yet classy: the future & big data
Today, we generate and store more data than ever before. And the amount of stored data is growing exponentially. Yet big data is not only about quantity. The quality of data and how it is evaluated also count. “The data itself is useless,” says Frank, responsible for big data at ING in Germany. “Its value depends on how you extract information from it and then use that knowledge.” Frank explains how to successfully harness big data – and the role it plays at ING.
“Data is the new oil.” ING expert Frank agrees with this oft-quoted comparison, as it underscores how fundamental big data is for digitalization. Like oil, data is a valuable raw material, but it is also the lubricant which keeps everything running in the complex mechanisms of our digital future. Patterns are recognized, conclusions are drawn, actions are triggered – automatically and at lightning speed. Consumers are already benefiting from this today. Big data facilitates a convenient online shopping experience. Recommendation engines suggest offers that truly interest us. We can track the delivery of packages live. And virtual assistants make everyday life easier at every turn. As Frank puts it, “Big data is certainly no magic bullet. But it opens up a whole new range of possibilities.”
A success story with a history
Data has always been at the center of the information age. It all began over a hundred years ago with punch cards, a storage medium that now seems antiquated – but it was machine-readable. A series of more powerful storage media followed: from magnetic tape to the floppy disk, hard disk and SSD. At the same time, the computing power of our processors has continued to increase exponentially. And so has the amount and type of data generated: from GPS information to market movements to text documents. A colorful mix of data now facilitates completely new services thanks to innovative approaches like machine learning – so even unstructured data can be used, in addition to the classic relational data of data warehouses. “Today, we’re trying to capture anything that can be measured. Things we can then evaluate: What does the user do, how long did they hesitate, which click came next?” explains Frank. To make use of these findings, near-time solutions are crucial. And that brings us to fast data.
Fast data: immediate, precise, efficient
Whether you look at it as a consumer or from a business perspective, when event data is evaluated in real time via fast data, you increase the benefit for all parties involved. “The art is to combine current data with historic data,” explains Frank. That is the only way to filter the sea of data to identify relevant patterns, anomalies and events. After analyzing a process, the results can be fed back directly into the process itself. For example, a customer interrupting a request process can trigger a call from an agent. Reports can be generated in real time. In parallel, social media trends can help adjust the marketing strategy. Big data is also extremely useful in IT administration, as Frank explains: “Based on recurring patterns, the machine can recognize, for example, that a certain error might reappear next week – or it can facilitate the early identification of complex error triggers. This not only leads to greater system stability, but also saves engineers a lot of analysis work. As a result, they can spend more time on new and ongoing developments.” However, Frank adds, “everything has its price.” Consumers themselves are becoming increasingly “transparent”. This price must be balanced out by providing more added value to the user. In the end, what counts is the transparency of the companies that rely on big data. If users know what is happening to their data, they can make an informed decision.
Big data in online banking
So what does this all mean for banking? Of course, there are countless use cases for big data. As an online bank, ING Deutschland is inherently well positioned. All processes are digital from the outset. No one struggles with legacy structures from the mainframe era. As a result, solutions can be implemented and scaled seamlessly – whether it’s marketing, individualization of the user experience, internal productivity tools or transaction management. “Big data is especially influential in the field of fraud detection,” explains Frank. How can fraudulent transactions be quickly detected among thousands of simultaneous payments – while still adhering to data protection and regulatory requirements? To detect suspicious patterns in the data stream of transactions in near-time, ING engineers created a stream data platform with an event-driven architecture. It is based on Apache Flink, a stream processing framework with multiple connectors. The “stateful computations” typical of stream analyses are performed in-memory with low latency. Applications can be containerized “as a service” based on the OpenShift infrastructure. The result is a solution with scalable application containers that allow IT teams to quickly exchange individual components – a highly future-proof approach.
The development continues: pioneering IT at ING
Fast data analytics is an innovative field with barely any standard solutions available. Even though ING is a digital challenger bank, industry-specific regulations set limits on the use of cloud service providers. In plain language, this means that ING's IT professionals must – and are allowed to – develop a great deal on their own. “We have done real pioneering work, particularly in our fast-data projects,” sums up Frank, such as in the use of Eventbus technology or NoSQL databases. “It is extremely important to remember that big data is not only about technology and data protection, it is also about mindset and culture,” emphasizes Frank, “and for us, it’s also a passion!” IT jobs at ING are attractive opportunities for ambitious IT professionals. The Frankfurt a.M. and Nuremberg locations are constantly on the lookout for new employees. You can expect an agile environment in interdisciplinary teams, plenty of room for your own career development and connection to the global IT community. But above all, you can look forward to exciting new challenges, and there’ll be no shortage of them in the future. Because when it comes to big data, once you've achieved something, you can never rest on your laurels. “It's important to keep at it. One day, everything we’re using now will also be outdated,” says Frank. If you want to stay on the ball in digital development, find out about the opportunities at ING on our Jobs page.