Reinventing banking in the new decade using AI

Our banking system is today facing increased competition. Innovative services are on the rise, thus making the traditional banking we have known for decades slowly obsolete. There seems to be a drive for more self-service round the clock banking services, which are cheaper and more convenient for the customers.

All of this is happening within a global scenario where compliance is becoming much more stringent, and security risks are increasing exponentially. Finally, with transactions happening at the speed of light, the volume of data is growing at record speed, thus making effective decision making difficult. Of course, all of these goals may sound daunting for traditional banks, but they can be easily achievable with the help of Artificial Intelligence (AI). Let us have a look at the technologies which are transforming traditional Banks into Fin-tech organisations.

Digital banking assistants offer users financial coaching and access to the various services provided by the bank. One can consider them as being a personalised cashier that knows the clients individually on a first name basis and is ready to give them all the information they desire at the click of a button. Such systems make use of Human Language Technologies (HLTs) which converse with the user using a natural language (such as English, Chinese, etc.). This technology increases the overall user experience drastically since the user interacts with the virtual assistant as if he is speaking with another human. Furthermore, the assistant is capable of upselling other banking products, thus helping the customer to explore new possibilities.

Client Profiling is another area where AI can assist. By going through their demographic and historical data, banks can automate the categorisation of clients based upon their risk profile. Advisors can then use this information to offer clients different financial products in an automated way and based upon their profile. The advantage of using such a system is that the client is not subjective to data-induced bias. Furthermore, since the AI is continuously updating the profile of the client, it can immediately notice any growing risks and alert the bank before it is too late.

Contract analysis is a somewhat repetitive task which takes a lot of resources. Because of this, Managers and advisors can assign this task to an AI assistant. Optical Character Recognition (OCR) can be used to digitise the documents (both typed and handwritten) and send them to the contract analyser. The Natural Language Processing (NLP) module will then interpret the text, analyse it and apply business rules. The business rules applied can vary; from straightforward ones (like checking the status of checkboxes) to extremely complex ones (like checking the logic behind the text). The AI model uses past contracts for training, thus increasing its accuracy drastically. Large companies like JP Morgan report that using such an approach; they managed to condense 360,000 hours yearly of work into a few seconds. Once the system completes the processing, it provides alerts with suggested corrections at very high speed which are reviewed by Managers or advisors.

Customer churn is another major headache in such a competitive environment. AI can be beneficial in predicting churn based upon the client’s interaction (or lack of it) with the bank. The system can quickly analyse all the data about a client, make a prediction and creates a prioritised list of clients who will probably terminate their business with the bank. The manager can then intervene directly with the clients in the hope of retaining them or at least, understand why they opted for a competing bank. Research shows that it is much more difficult and expensive to acquire a new customer than retaining an existent one. Mass marketing fails in these situations because customers are so diverse that the message will not get through. Thus, customer analysis and churn predictions are vital to creating marketing initiatives specific to particular customers.

Algorithmic Trading uses AI techniques to trade shares, stocks, etc. It achieves this by analysing data coming from different sources such as past historical data, news items, etc. and taking incredibly fast decisions. The patterns found in this data is complicated to spot by a human, or it is physically impossible for a person to process that tsunami of information promptly. Of course, a computer can automatically execute trades rather quickly and secure a sale or a purchase at lightning speed.

Automated Research tools handle the background research necessary to approve or decline financial investments. This task is extremely time-consuming for a human to do, but for an AI, it is rather simple. The system is capable of performing sentiment analysis on information about individuals or companies. It can detect if people have a negative or positive sentiment about the entity and thus provide managers with critical information. Advanced language processing techniques can help researchers get a quick overview of a company’s financial reports by zooming-in on crucial topics and automatically highlighting specific passages.

Automated Valuations of investments is another application of AI. Using various features such as age, historical values, etc.; the system can calculate the worth of an asset. Traditionally, such a system was initially used in real-estate to calculate the valuation of houses. However, financial firms adopted similar approaches by looking at economic indicators, growth predictors, etc. to reach a final estimate.

These were just a few examples of where the banking sector is using AI. In the coming years, we expect to see many more innovations dealing with the operations, marketing, sales, customer experience, revenues and quality of deals amongst others.

AI is a tool that is changing the way we bank forever, and it is ushering the era of modern banking. Our banks cannot stand still, or they will die. As the old Chinese proverb goes “Be not afraid of growing slowly; be afraid only of standing still.”