Customer Ignite Talk: Maurizio Poletto (Chief Platform Officer & COO, Erste Group) & OpenAI

[music] >> Good afternoon, everyone. My name is Fola Odundele. I lead part of our enterprise business here at Open AI in London. Um and I am joined by Maurizio Poletto. Maurizio, they've given me a pretty impressive bio, so I'm going to try and do it justice here. You are the Chief Platform Officer and CEO of Erste Group with a background spanning design, innovation, and fintech. Uh you've been a driving force behind transforming digital banking at scale, most notably through the development of George, one of Europe's leading digital banking platforms. And really, today's conversation is really focusing on your adoption of AI across the business. Um I personally enjoy these chats because they are the real-life examples of customers that go through challenges, have those learnings, and we're here to share those learnings today with you all. So, Maurizio, Erste has been a digital banking leader for a number of years, but something has changed in the last 12-18 months. Um what makes AI that board level platform level priority now? >> On the board level, the excitement is there from since day one. The difference is that on day one, um the excitement was all ourselves reading and experiencing this by ourselves privately and having a bunch of PowerPoint presentation. That's where it ends. Now is same excitement, even more, with a bunch of problems to solve. How to roll it out to your employees, how to make them using it, uh how to make it legal, how to make it compliant, how to convince uh security people to open up and not to close. So, the excitement is the same. It's just that it follow up with a list of challenges, and I think that's what I like the most because without solving those challenges, uh you don't generate value, right? And I think we are right in the middle of that thing. So, um we are we are still super excited on the board and we have always been. Um the difference with AI is that this is not an hype and is here to stay. And I remember with blockchain there was a similar there was a similar situation 5-6 years ago, you may remember. Super excitement about blockchain, every problem in the world will be solved by blockchain, but it was never followed up by a list of challenges to to to to solve. Here is different. This is really really really substantially different and and and we are in the middle of that project or that process of implementing it. >> Excellent. And we'd love to hear more about that process and the challenges that you're facing. Um you're operating obviously in a regulated environment and as a regulated bank, how do you decide when to go fast, when to go slow when you think about things like governance, data residency, compliance, auditability, many many other things? >> So, in banking I can speak for many colleagues here. You have to start slowing down once you start accessing customer data. That's the moment where you start sit down and say, "Ah, I cannot bootstrap it anymore. I need to do it properly." Uh because probably that's the most sensible piece of information we have in our system. We are an industry built on trust and if you look at trust beside the fact that it's quite expensive, uh uh uh it takes you a while to regain that trust. So, as soon as you touch customer data or you want to start processing customer data, you have to slow down and think about doing it properly. Uh we have made the decision at the very beginning when we started to engage with AI about 2 and 1/2 3 years ago, uh that we didn't want many banks did a smart approach. They started with data which is doesn't have any customer data. They started with help center, general information, data which is on the public portal, easy. We for several reason we started the hard way. We immediately started to connect customer data right away from day one. Lots of mistakes in that process as you can imagine, but we went the hard way because we said eventually this needs to happen, so why don't we do it right away? And and there therefore there you need to do it right. You need to have compliance in place, you need to have security in place, you need to make sure that things are done properly. We need to make sure that everything is done in a correct way. And yeah, it's a bit it took us a bit it puts a bit of gray air on on our head, but I think it pays off. We have a certain amount of knowledge on that topic right now that give us the confidence we can move forward faster. >> Great. And one of the areas that you started first is is rolling out chat GPT internally and I'm I'm curious, what have you learned from rolling out chat GPT AI tools that has changed how you think about broader employee adoption? >> So, in the group we had a very simple strategy. Um trivial. I'm in charge of the group, so I'm not running a specific country. I'm on the holding level and I'm in charge of different country and we said for productivity tools every country can almost do whatever they want. So, go on fast because we knew that I don't want to become a bottleneck for my colleagues in the country to do, I don't know, chatbot that help them up their team to uh learn about document, working instruction, policy and stuff like this. I say, on those stuff these are the tools, go on as fast as possible, learn as much as possible. On customer facing, we had a clear rule and say you don't touch customer facing. We do this centrally. Uh because in the group we have a one digital platform across all the country, and we wanted to do customer-facing AI conversation and advisory and support on the app with customer data. We want to do it centrally, and so that we we keep ourselves restricted. And, um, so this was the overall strategy. What we found out is that not all our customer ask the same question we will ask. And so, uh, we also learned that many of our retail customer, we started with retail, of course, are not ready to have a conversation with the bank. They're ready to come into the branch and have a conversation with a person, but they're not they're using our digital banking app as a tool, as a service. And they takes a bit of time for them to understand that they could they could have a companion where they can have a conversation with. This is a process that will take a while, and it's still taking today. And so, for instance, one of the nudge I can tell you is every time we implemented AI, we always had two way of doing AI. We have the reactive AI, means the customer open the conversation and start talking, and we have the proactive AI, when we push a conversation nudges on our customer and see how they reacted. And as you can imagine, the great result we got it on the proactive one, when we push, when we give our customer an idea which kind of conversation they can have. Because the reality is very simple. Many of our customers simply like AI, but have no idea what to ask. They just don't know. So at the moment we have in this strategy that we about counter balancing proactiveness and reactiveness in a in a way that we get this new technology and those conversation familiar on our customer level. And we are in that process right now. >> Amazing, and I know that the George app and digital experience that you have is is used by millions of customers. >> Yeah. >> And I'm curious and maybe the audience are curious, what should something like George look and feel like? So, you talked a lot there about the proactivity. Completely agree, but should it be more self-service, a financial coach, something else? >> I I don't know. If I would know, I probably I don't know. Uh I don't know if I have a very simple joke I do in the company. I always say, "Oh, I remember for- forget that. Uh you remember when Alexa came into place? Conversation. You remember that? And at that time, I remember lots of consultant and company came to us and say, "People will not talk to you will not use your app. They will ask questions." Uh because now with Alexa, you can ask question. And and I said, "Listen, I'm coming to the office with the public transportation every day. And what the first time I see someone asking their phone, 'Okay, what's my balance?' When I see that happening, then I will probably start changing my mind. I I I don't see this yet. Uh And I remember this crazy experience someone was showing me that you were going home in this one of these promotional video from consultant, you go home and you say, "Hey Alexa, uh what is the last transaction um I I I happened on my account?" And then Alexa said, "No problem. Open Sign up on your app. Take your app, sign up on the app, and then I can give you that information." And I said, "Well, that's uh something wrong in this, right?" And so, um I don't know if banking will be conversational. It may be. I don't know. Um I still believe that doing doing a transaction, if you have build your interface right, is much easier to do it by clicking some buttons and putting some number than telling someone to do a transaction. I'm not ruling out that we will get there and say, "Do Send the money somewhere." And the system will do it. But, that's not my challenge. My challenge is that or our challenge is that at the moment we are roughly providing financial advice to roughly 20% of our customer base. Which is the customer base that decide to come into our branches. Um and I think our ambition is to provide financial advice, help them to really take care about their finances to the other 80% of our customers that at the moment don't do this. They're just using our tool very functionally to make transfer, to see where they stand, but they don't really get the advice. That's a big challenge. Because if you analyze this missing 80% are most probably the people who need that advice the most. Because normally the 20% that get that advice today are normally a bit wealthier, a bit older. They have time to come to the branches. They have and normally we serve them great. How can we serve the other 80% to really make an impact on their life? Our vision in the group is that we are here taking care about like a doctor taking care about probably the second most important thing in in customer life in in people's life, which is their money, right? There is health first, then comes the money. Someone said number two is love. Someone said maybe money comes before love. But, whatever, but either the second or the third most important thing. And our ambition is to do anything we can to serve all our customer and not just the wealthy one. And I think AI comes very handy here. >> Yeah, and I think democratizing access to the tools and capabilities for everyone, uh we certainly align with that. Um on that note then, what role should partners like an open AI play in a bank's AI transformation? Technology provider, co-builder, strategic advisor, all three? >> Mhm. Um in our preparation call I was mentioned one thing to Folay, and I said you guys from OpenAI put that in a very bad situation. Why? Because you have very successful consumer product. I'm one of your pro user. My best $23 I ever spent a month. It's uh but you also set a standard. You set the standard. And so, everything I deliver to my customer with all the restriction of my industry, my risk department, my security department has to reach that standard. No one will say, "Yeah, your your AI conversation is a bit worse than what I get in OpenAI, but I'm expecting it because I'm accepting it because you are a bank." No one will do this. They say, "Well, that's OpenAI can do it better." And so, you you set that standard, you need to help me to get to that standard. That's what we need from you. >> Challenge accepted, Moray. Um and absolutely, that's why we provide the tools, the capabilities, and we definitely want to continue to >> It's like about I told you, it's like about Amazon make everyone use that everything can deliver in one day. >> Yeah. >> You know, at the beginning with Amazon, this was a money loss system, right? For many years they were losing money with one-day delivery. But everyone has has to catch up. Uh it's a bit similar what what you are doing and some of your competitor do it too. So, the the the bar is really high, and there is no uh you know, there is no uh there is no space for a partially good experience. Either it's good or it's not good. >> Customer expectations they're evolving and changing. Absolutely agree. Um I can see the time we have time for one last question. Clearly, this is a space you're passionate about, something you think about a lot. You're on journey and then making progress on that journey. But to those in the audience who are just starting out working regulated financial industry, what's a piece of advice you could give them starting in the early phases? >> I I I start to give advice, but what I can tell you is I we accepted to make mistake. So, we accepted So, now I I think we are building this AI platform our AI. So, we have a digital banking platform and the AI is a platform within the platform. And we are already since 2 years we already at the second version of it. Not a bit of a version, a bit different. I So, the first version we have to redo it completely. And I'm not ruling out that the third version we're going to do in couple of years needs to be completely different. So, you have to accept that this is this is the case. It's like like every little startup you start with a monolithic application then once you are there you figure out that this is not scalable and then you need to do it differently. And you need to take the whole thing and re re reconsider it. And um in some company this is considered in some organization this may be considered a waste of money. In in in my organization in our organization we considered it as the way to do it. And and I'm already planning that I need space in probably a year and a half to to redo it again. The game is about doing it faster and hopefully with some coding assistant and so on we can redo it completely faster. But yeah, you have to accept that you have to do it wrong couple of times before you do it right. That's that's the only thing I've learned and I don't know if it will help any one of you to do that, but >> No, I think that's great advice. Mari, thank you so much for this conversation. >> Thank you Thank you very much. >> [applause]