Accenturen Nordic Future Tech Lead Hans Kristian Aas oli haastateltavanamme alkukeväällä Suomessa käydessään. Kysyimme häneltä, mitä kaikkea finanssialan yritysten tulisi ottaa huomioon tekoälyä hyödyntäessään. Yksi Hans Kristianin tärkeimmistä viesteistä oli se, että pelkkää prosessien automatisointia ei tekoälyn avulla kannata lähteä toteuttamaan, vaan tavoitteena tulee alusta asti olla uusien, tehokkaampien ja asiakaslähtöisempien prosessien keksiminen. Lue Hans Kristianin haastattelu (englanniksi) tästä.
What kind of AI applications companies are now developing in financial services sector?
I think most of the banks are starting off with their operations and their back-office, because that is where most of the money is. So, they start off by assessing the different processes and then see how AI can be applied to automate those processes. What you typically discover then is that fixing the processes in the back-office often ends up with automating bad processes. The leading banks are looking at this more holistically and are assessing the whole value chain and the whole customer journey. Then reinventing the process, coming up with the dream process for the customers and then applying different technologies. That’s a very different way of doing it instead of just automating the processes as they are today.
What are the most common pitfalls in utilizing AI?
I think one of the most common pitfalls is to say that now we have this wonderful new technology and AI, let’s try to use that on everything. That is usually not how you end up with the best cases for AI. So, taking a more strategic approach and really understanding the customers, then applying the AI to do processes in another way. What we see is that a lot of processes today have a human judgement in them and they use unstructured data. Those are the two things that you don’t want when you automate a process using traditional tools. By applying AI, we can actually structure the unstructured data so we can pick out everything that we want from a free text and we can replace human judgement by using machine learning.
What are the success criteria in using AI powered solutions in financial services?
The success criteria are thinking through the entire value chain, taking on the whole customer journey, use that as a basis for your new processes and make sure to use your whole toolkit of process improvement. Just stop doing things is quite efficient. Then add into the new AI technologies as virtual agent, as machine learning, text analytics, speech and machine vision. Using that whole toolkit and using the right tools on the right processes, you will end-up with much better results.
We see, for example, that AI driven virtual agents have been adopted very widely. Virtual agents are robots that interact directly with customers. Now also, machine learning and text analytics have the ability to understand the Nordic languages, so that is opening up a lot of opportunities. We are estimating between four to eight weeks to teach a virtual agent a new language.
What about responsible AI?
Using algorithms is a bit different than traditional IT. A new thing that we have come across is that for every problem that we are trying to solve there are usually two to three different algorithms to choose from. We see that assessing those algorithms becomes more important. We have some algorithms that are more like a black box, so it’s very hard even for skilled people to understand how the machine is actually making a decision. Then you have other algorithms that are more transparent, and I think for the financial industry, choosing transparent algorithms which you can understand how they make the decision, could be a better choice than going for the black box.
How can Accenture help its clients with utilizing AI?
We’re experiencing that Accenture has a unique position into AI. We have the technology in our backbone and AI is also a part of our DNA. Combining that with the industry insights and new creative methods as supplying design thinking methodologies. We see that we can make completely new services that makes the banks more competitive and more efficient on the inside.
Addition to utilizing AI, what is it that you consider another critical issue for banks at the moment?
The Nordic banks need to balance and for me it’s like they are having two different shoes on. You have the business shoe and you have the FinTech shoe. The business shoe is, of course, very important because you need to reduce operational cost and you need to be more effective. At the same time the whole industry is being disrupted by new companies called FinTechs or other industries like Telcos and technology companies like Facebook. At the same time banks need to have that FinTech shoe on, reinvent their processes and find new ways to deal with the competition. That balance is, of course, not easy for a large company to do, but that is where the banks need to succeed in order to survive.
Katso Hans Kristianin vastaukset myös seuraaviin kysymyksiin videolta:
Hans Kristian Aas,
Nordic Future Tech Lead,