Automation using AI in banking and finance

Artificial Intelligence (AI) tools for the Financial Technology (FinTech) sector, and all areas of Banking, are getting quite abundant—overwhelming, in fact. Interestingly, as it turns out, a lot of it is hyperbole. They are not to be ignored or overlooked of course, since many are incredibly useful, and are vital for what they offer to the FinTech sector. We must select appropriate tools.
We need these new tools to be competitive, and our customers expect us to have them. They grant small, boutique style banks the same powers and capabilities as those of the centuries-old massive institutions. Naturally, there are some distinct advantages conferred with billions of dollars in resources for the larger institutions, but that also creates a lot of inertia which limits the ability to change.
The new competition is agile, flexible, and willing to embrace changes in very short timeframes. And, as we know, customers are fickle; they’re willing to change providers for a perceived advantage, or a level of responsiveness missing from their former institution.

What Customers Want

We are currently dealing with more customers per office, and per employee, than ever before. Our audience is growing, but older, traditional methods of customer service can’t keep up with the demand nowadays. We need automation, now more than ever before.
Old Mr. Fezziwig, Ebenezer Scrooge’s employer, went out of business because he clung so tenaciously to traditional methods. His heart may have been in the right place, but when you can no longer serve your customers properly, it is time to review your methodology!


Seventy-five percent of customers surveyed want ChatBots! That is not a mistake, though it might more accurately be phrased as “Customers want an instant communication interface, with no wait-time, and usable answers.”
Research has revealed that the pace of our life has increased so much in the last couple of decades that human attention spans have dropped to less than a minute unless stimulated continuously. Web surfers will abandon a site if its homepage remains blank for more than 5 seconds!
Studies have also revealed that people waiting in a telephone queue (“your call will be answered in priority sequence”) get irritated after 17 seconds, and that continued waiting quickly escalates into anger and rage. That is one of the worst possible ways to treat a customer because, to their eyes, it shows that you don’t value them.
On the other hand, when customers know they will be faced with a wait, they prefer to use a Chat Platform with their Smart device because it is fast, responds instantly, can handle most of their inquiries accurately, and perform most of the typical banking functions which are available by phone. In many cases, when asked, customers are not even sure if they were chatting with a person or an automated system.
Some new deployments use voice-recognition technology to actually “speak” to clients, rather than using a typed interface. Most do this utilizing a translation engine, such as Google (which currently supports 100+ languages, translates images of 37 languages, and manages “conversation mode” in 32 languages, in a smartphone app.
It won’t be long until anything said, in any language which can be translated into or out of English, will be completely understandable. You only have to look at the success of Alexa, Cortana, Lyra, or the other AI precursors (they’re not AI yet) to see how far we have already advanced.

Getting to know you

More importantly, callers are generally identified by their incoming phone number, and there is a record of previous interactions with a customer so that there is no need to start anew every time they call the bank.

“Hi. This is Mr. Smith, and I was calling because we are waiting on financing approval for the new cottage. It closes in 15 days…”
“Yes, Mr. Smith. The mortgage has been approved. You should have received a copy at your registered e-mail. It was sent at 8:26 AM.”
“Yeah, I’m on my phone and can’t get my e-mail right now. Can you send a copy to my phone?”
“I just did. It should be there now. Is there anything else I can help you with today?
“No thanks, that’s great. Goodbye”
Whether as a typed conversation or using the latest speech technology, this represents a thousand-fold improvement over “your call will be answered in priority sequence.” And the more frequently they call, the greater the understanding that will be developed about this client.
And, of course, if they ever need to speak to a human customer service representative, the AI can connect them to a staff member who has a similar personality or disposition. This increases the likelihood that the customer will leave with a memory of a situation quickly resolved by a knowledgeable worker.


We already possess an unmanageable amount of data. Getting it from point “A” to point “B” can be a massive undertaking. Old-style copper networks used to be able to handle the demands of moving our data, but the sheer volume of information combined with aging infrastructure and the physical limitations of hardware create clogs and backlogs.
Unfortunately, employee-to-customer conversations often go like this:
I’ve requested the file, but the system is really slow today. While we’re waiting, how about we look at [something] again and make sure it will suit your needs?
…which is ridiculous, of course. The employee is left to flounder while the “slow computer” fetches the file.
Modern AI systems drive an enhanced system called Network Function Virtualization (NFV) which runs on a Software Defined Network (SDN). The AI acts as an overseer for all network traffic and creates pathways for data to travel, sending information over underused circuits, and freeing space elsewhere.
This system significantly enhances the speed, efficiency, and integrity of data, minimizing complaints about these types of issues. It makes a great intermediate step until you can upgrade to an optical network (if necessary) or make other changes because it uses your existing hardware and minimizes capital expenditure.
A self-maintaining network pays for itself many times over. Happy customers are their own reward, and you will have to deal with fewer complaints from your workers. These benefits are of incalculable value to any business.

Irrefutable Identification

The idea that somebody can pass an RF reader near a purse, wallet, or your pocket and steal Personally Identifiable Information (PII) in just a split second is a bit of an intellectual leap. Some folks think it is amusing to hold their wallet near a retailer’s pay-point when making a purchase, and leaving it up to chance as to which card is selected to pay for the purchase.
We need to be more careful with our PII. Banks might be wise to give away RF-proof wallets as a free premium to their customers. It will prevent many losses.

Identity Analytics on-the-go

Tap & Pay is not going to go away anytime soon, but AI will help us by making sure that the person with the card is the one that is supposed to have it by adding 2FA (Two-Factor Authentication). Possessing a PIN is one solution, but adding a signature pad is better. The way you scribble your name, where you dwell when writing, how much pressure you apply to the pen making certain letters, all these biometric features uniquely identify you and you alone.

Identity Analytics Online

Even more useful, an AI can reliably identify a person with biometrics. This will eliminate the need for passwords. The AI will know their typing-tempo, common misspellings, and words where they typically hesitate or dwell. It will even be able to distinguish between a photo and a real person by recognizing the rise and fall of shoulders and chest as someone breathes, along with the blinking and movement of eyes.

Internal security

Beyond better customer service, AI-driven biometric identification means that authorized people can get into restricted access areas without the need for keys, codes, or passwords. They can perform specific actions, and there is consequent record care for every activity.

AIs can also recognize unusual situations. It could be trained to identify the actions that precede a bank robbery, for example, alerting staff, security and the police department before the event even takes place.

Constant Reporting = Loss Prevention

Forensic Accountants can show you where a project went off the rails; where a company started to hemorrhage money because of a profound error. The problem is that they are not called in until the problem becomes very obvious, often beyond the point where it can be fixed.
You shouldn’t have a quarter of a billion dollar write-down because it was not identified until the Quarterly Report. Artificial Intelligence recognizes problems much more quickly than humans do, and action can be taken immediately. Even catching a series of small issues can save you millions of dollars per year.

Not only does it prevent losses, but it provides new levels of accuracy, and virtually eliminates the immense workload generated at the end of each quarter. AI will allow you to get a completely current overview of your organization, right up to the moment you read the report.

Automation of Ordinary Functions

People perform drudge work because it must be done. It’s not enjoyable; it’s not rewarding; it doesn’t often increase one’s sense of self-worth. Worse yet, it takes time away from other work that might be all the things which it is not!
Automating mind-numbing tasks doesn’t mean letting people go, or decreasing staff. Instead, it is an opportunity to put untapped creativity to work with new ideas, fresh outlooks, and innovation. Staffers, crushed under the weight of the mundane, could turn out to be the best idea factories you’ve ever seen.
Improving the Customer Experience

Predicting the Future

Understanding how customers use the resources of the bank is child’s play for an Artificial Intelligence. While the customer can never know all the services offered, the AI does. It can create a custom package that fulfills all their needs, without any extras, thus saving them money, and inspiring long-term loyalty to the bank. They are happy, and happily recommend the bank to friends, which is better than free advertising.
Truthfully, the happiest customer is the one who receives offers precisely when they need them the most. Predicting the future may seem like an arcane art, but an AI can see developing situations. Knowing that the customer rents or purchases a new car every three years allows early notification of financing options. That goes for anything they do on a regular basis. This foretelling of the future is precisely the same thing that Netflix, YouTube, and Google have mastered already, so they can present the next thing you are most likely to want to see.


The Federal Reserve, the FBI, and even the CIA will be pleased if you can spot money laundering here and abroad. Unfortunately, it slips by mere human observers because crooks are so clever.

An AI programmed with forensic knowledge could spot something suspicious immediately, meaning that you won’t be faced with fines for negligence. Better yet, you get to be the hero of the story, gaining favored status with those agencies, and the IRS might even be grateful for the additional revenue. This is a winning scenario from every perspective (except the criminals themselves).

The Takeaway

Once we teach an AI to analyze and integrate all of our data we can free employees to do more creative or essential work. Whether it’s answering common questions for customers as a ChatBot, or finding fundamental flaws in investment strategies, AI is the grease for our economic engine. It will make everything run better, faster, and at a reduced cost.

This is a significant change in the world of business. If you want to learn more, we provide webinars to show you where the future is going and how you can get there. Don’t shortchange yourself. Join us today!