Exploring the Future of Mortgages with AI and Digital Innovation. With Geert Van Kerckhoven from Oper Credits.

Speaker A: Foreign.

Speaker B: Jeans presents Voices in Payments Expert insights for a digital payments world.

Speaker A: Welcome to our podcast. Today we have Geert from, uh, from ope. Um, Geert, would you like to do a short introduction about yourself, uh, and um, and oper?

Speaker B: Sure. I am, uh, Gersten Kirikova. Uh, I'm co founder and CEO of operations. Um, what do we do with Oper? We basically have a white, uh, labeled SaaS solution to help banks to digitize their mortgage process. So mortgage origination from contact to contract. Uh, and we do that in continental Europe today in six markets with roughly 20 lenders.

Speaker A: Nice. Interesting. I remember when doing the mortgages myself as an entrepreneur, horrible, uh, horrible process. Uh, almost, ah, stopped doing it, but it ended up. It worked for me. But I can imagine, um, a lot of people feel this pain point. Can you maybe share with, uh, us why did you start there? Maybe start with your background a little bit.

Speaker B: I come from the technology and the banking side now. I've also done two mortgages in my life and I really emphasize that, uh, it is a very big decision and not every decision is the same. So it's a very tough thing to go through, um, and very complex. Um, but actually my career started, uh, 15 years ago. Uh, I joined a startup and they built actually core infrastructure for lending. Um, and back in the days I just wanted to join a startup. Uh, so I thought it was. Personally I really love building, so I thought it was exciting. Um, I had no clue what I was going into. And basically they had a first client, they had a PowerPoint, a pitch deck and a Word document. And it was like, now we're going to build it.

Speaker A: Interesting. More as an entrepreneur about setting your own company. What is the biggest learning that you had in the last, uh, five years?

Speaker B: To be honest, maybe my most important learning is, um, that I think I found my true passion. Um, and I think the second thing that I really tried to explain to earlier founders, and you also alluded to it with Paying Jeans earlier, is how crucial talent is. And that is something that as an entrepreneur, everybody tells you, be, um, very picky in the people you work with. Um, really use the, I mean, really take your time to find people that click cultural fit. Um, and you hear that in every book, every podcast, every founder talk, and still you violate the rule every week because you want to go fast, because you're hungry, because you're a natural optimist and you're like, I, you know, brahm is nice, but I have a few caveats. And if you have a caveat don't. If there's doubt, there is no doubt. But that's a very hard thing because I feel, I don't know if you also have that experience, but that's my key learning.

Speaker A: Amen I would say. And I can imagine you need that passion as well if you want to change, um, or drive change within banks and financial institutions which are, I think you need a long breath in order to really change that. Can you describe the current landscape, uh, of mortgages digitalization?

Speaker B: Yeah. So if you look at the current mortgage industry and I'll focus on Europe to keep it easy, um, I mean today still a lot of the mortgages, I mean it's a very paper based process today. So if you see across the large markets today, the majority is still the majority of the let's say origination journey or the sales journey or the you know, starts in a branch or starts at an intermediary depending on which market, at a broker. And then there's a really large paper trail that moves up to the underwriting decision. What you've seen post financial crisis is that and also through a lot of regulations came to protect the market, to protect the client. And if you see how that has been solved, well a lot of red tape has been stuck to it. So you just see that actually the paper trail has only increased. But, and that actually there, I think that was maybe a good thing coming out of the pandemic. A certain point in time the paper trail in uh, a lockdown scenario, paper trails became very, very tricky. So we've seen that around E signatures around automated kyc, around let's say open finance and open payments like solutions have popped up. Also property appraisals um, to make sure that you can digitally know is the apartment really worth €3,400,000? I think there we have seen the last two to three years. I mean adoption. The only thing what we learn now that's a challenge in the industry is that we go to a bank and a bank says yeah, we have electronic property evaluations with service A, we have, we have an open banking API, but we're only using this for 5% of our production. And it's all on the side of our systems because we have a mainframe that runs in branches. And so today I think a key challenge is there is a lot of nice innovations around but it's a patchwork, um, and not connected.

Speaker A: Yeah. So can I phrase it? Am I phrasing it correctly? If you say that technology in the pandemic and technologies have helped uh, increasing awareness of the possibilities but also now limiting uh, the potential use of it uh m. Because um, of certain complexities within the technology because of the different technologies that are not collaborating both 100%.

Speaker B: I think what it's more of a lot of banks are in the transformation of um let's say transforming their core. It can be core banking systems, can be their core sales systems and I um, think those innovations will only really bring a return on investment when they are interconnected.

Speaker A: Talking about AI, how do you think or how already maybe already it has uh shaped uh the mortgage industry specifically on automation and loan approvals.

Speaker B: I'm honestly quite excited. Why especially 5, 6 years ago I was already working um in a different profession on automated credit scoring, automated loan approvals using machine learning. The only tricky thing with machine learning was that you needed to trade on your existing data and then you came into the, I mean then you need to come to discussion where are we getting the data? We need to get it out of an old mainframe system. It needs to be cleaned up and before you've done all of that basically the project would be gone, you would have spent millions. So there was already on machine learning a lot of possibilities and I think we have some good use cases there. I think now with the foundation models with let's say Genai and LLMs, what LLMs are inherently good at is predicting words and interpreting text. And if you look at a mortgage process it is one text based problem statement. I mean we are verifying you apply for a loan, uh, we need 50 data points from you. Uh you'll probably give them to us but then you also give between 20 to in some countries 50 to 60 PDF documents scans to verify if everything you said is actually true. Um that's a lot of text and I think foundation models are inherently good. If you put the right prompts on top of them, you ask the right questions, are inherently good in supporting that process.

Speaker A: And that's kind of looking at specific parts of AI. AI is a kind of a general term uh obviously but that's, that's for the upcoming years. And, and, and as you say we still need to train those models in order to be really productive and really reliable. What is, would be the, has been the biggest contribution of AI or hasn't there been any so far that we can really say that that you mean in general in industry or in the mortgage industry?

Speaker B: Yeah, I think today the uh, especially on credit scoring, I think that's where it's used the most is you get the clean data in and you basically say, should we underwrite or should we, should we approve this, yes or no? I think there we've seen the best use cases. On the other hand, I also think the business rules engines there are also working. So it's more of a complementary service for when you have, let's say, more edge cases. But I think there, I've seen in the mortgage industry the best use case today.

Speaker A: And while thinking about all this efficiency, efficiency gains and the accuracy. Yeah, I think, uh, what, what the kind of, the impact is for the consumer or for the customer.

Speaker B: I think the easiest one that we use is that we want to get to 81% faster mortgage decision. That's what it's all about. It's about certainty. It's about you seeing a property, especially in heated house market, you're seeing a property and knowing, can we do it? And, uh, and can we do this? And of course, why doesn't a bank give you a black or white answer? That's because, yeah, the. Normally your situation is always nuanced. You know, if you're an entrepreneur yourself, there's always, there are always, uh, edge cases. So we're saying for 100% of the cases, we want to get 80% or 80 or 1% faster to a mortgage decision because that saves time for you as a borrower, saves time for your loan advisor, and saves time ultimately for people in the back office doing the

Speaker A: staring, compare and also creating more transparency maybe along the way.

Speaker B: Yeah, that's, that's on the product decisioning. Absolutely. Um, but yeah, transparency, I think transparency in giving you certainty that you can do it. Absolutely. Transparency on what is the exact risk policy is something that is not always that easy to share directly with a client.

Speaker A: You already mentioned open banking in the mortgage industry. Open banking is a. Yeah, it's not something that is here for already a couple of years. Can you explain a little bit how that has impacted the mortgage industry so far?

Speaker B: So for us, what we like about open banking is the fact that we can have access to mega companies that access to accounts, I think is key. Why? Because most of the time your salary is deposited on a bank account and instead of checking five documents or even getting a PDF of your account statements, which I think is crazy. Um, I mean, open banking gives us a lot of data of, you know, what is the credit worthiness of a client. Secondly to that, especially in the, in the dark region here in Europe, you still have to submit something called a household, uh, account. Uh, so it's much more detailed than we know it here in Benelux, uh, where you basically need to say hey, we give so much money to utilities, we give so much money to a car. It's really a detailed statement of what is the family spending money on. Which today in some Austrian banks is still filled out paper pen, um there for an Austrian bank today we use open bank. So if you're a client of the bank and you have the primary account there, we just basically load it in, we do the mapping and basically we get a real overview because again if you're filling it out in paper you might be a bit more optimistic if you're really excited about that apartment in Vienna.

Speaker A: And what do we need to overcome these challenges?

Speaker B: I think adoption, um, also from our tests, um, sometimes we have in our, because we have a borrower facing uh, side in our product as well. So borrowers can also directly interact, I mean using our technology with the bank. So first and foremost sometimes clients, uh, or borrowers don't want to connect their bank account. That's uh, if you're a client of the bank it's fine like ING client will always do it. But if you are going to ING and you are with another bank, you think oh I'm not going to keep that so I'd rather upload documents. That's one. And secondly I think is adoption because adoption will drive the fact that we get cleaner APIs, we get better classifications because there still in some countries we connect to the access to accounts gateway. Uh we try to see if do we find income and then we don't find income. And again if we don't find a clean income we cannot use it in our own application and we're going to go back to salary slips. I think while for example in the uk, the UK is quite common. I mean in UK is really advanced on that. But again they had a very good implementation of PSD2. Open banking was really pushed forward. You then also the vendors in that space could connect to cleaner APIs banks start using it. So you get more traffic, you get also more complaints if it doesn't clean. And it's a flywheel.

Speaker A: Right.

Speaker B: And I feel on this side of the, of the river, I have this side of the sea, uh, we're not that far yet.

Speaker A: So it also has to do then with um, consumer expectations. Uh you say in the UK it's almost integrated in the consumer uh expectation as it is, you will be unhappy if it doesn't work. Um, so how are borrowers expectations changing now? And also in the rest of Europe when It comes to mortgage and payment

Speaker B: processing, I think, I mean they, they all start. These days you mostly start online. Why? Because you also start looking for a property online. But what I've seen happening in the last one and a half years is that, is that actually you want to start online, you want to play around, but then you want to speak to somebody as soon as possible. Because a lot of mortgage cases have. There's always a little bit of a. In Dutch we say there is a, there's a bit of a corner, there's a bit of a corner to it, special corner to it. I think it's something Flemish. Um, and so it's always specifically about the case. And so that actually today borrowers expect good advice. And that today is I think very, very tricky is to talk to somebody that is really knowledgeable and can talk you through your project. Um, I had a workshop with a bank yesterday about this that also said, yeah, during the interest rates that went up, they had so little volume that they reduced their mortgage advisor capacity. Now interest rates are going down again, so volumes are going up. And they say, oh, actually we don't have advisors anymore. So we actually really strumming to give good advice. So this whole paradigm of a client is going to do everything online when they buy a property for the first and the second time. I don't think that's a valid assumption anymore. So. Meaning that borrowers expect to speak to somebody when and then they got good advice, then things should move fast and digital again.

Speaker A: So we've discussed open banking, we've discussed um, the role of AI, if not a term fintech financial technology and um, upcoming potential possibilities within the fintech market. How would you think can that cross over with the mortgage industry? How can it benefit, how can it increase? Uh, are there kind of solutions that you feel like uh, outside of AI, outside of the potential with open banking that can increase the customer experience?

Speaker B: Yeah. What I think is something that I've been very optimistic about for only a few years. This is something you hear on every big conference if a banker comes up. But they're always like a borrower wants to buy a house, they don't want a mortgage. Like the killer statement. Right. But today, where in Europe do we have an integrated experience?

Speaker A: I cannot name any.

Speaker B: Exactly. I made a, I mean in that five star experience that I said earlier when we started oper, we actually said like, look, I'm browsing apartments in Amsterdam, technology wise, I want to know, uh, this apartment is affordable to me, this is not affordable by the Way I still have a studio somewhere in, in uh Utrecht which I might be able to sell. What's it, what's it worth? How does it go into my. So this full. Yeah, I'm going to, I don't want to do a buzzword bingo but this embedded experience, I think it's a uh, it's a no brainer but it doesn't exist and I think that can really create different borrower experience. Um but today uh, it is really Chinese walls between let's say the prop tech world because it's more proptech and the finance in it which by the way in payments um if you look at E commerce it is fully integrated also. Buy now, pay later. There you have this very nice um, let's say embedding but in the mortgage industry it's far gone and there I see a lot of potential to be

Speaker A: a leader who starts approaching those proptech firms. Right. Or and kind of make the transition happen. So you need to kind of look for connecting industries to, to drive change.

Speaker B: Yeah, correct, exactly. And I think everybody on the real estate side wants to go into that direction. But the what I learned as well but I mean maybe listeners can, can also uh build the changes here but I mean is that proptech people don't understand that much of financing. So it's two different worlds and bringing those worlds together is always very challenging. I mean always there's something that you know, how is the remuneration model? How is it going to work? What does the regulator allow? Um and then often there it halts and it gets back into a very simple lead model where I bring you a lead and then the bank takes over.

Speaker A: One of the topics that we haven't touched about is more kind of the regulatory part compliance and security. What are the challenges? How is overcoming.

Speaker B: It starts with the European, I think the European laws and regulations around, you know how do we, you know for example uh, are we doing digital advice for a client? Is that in line with the European and the local regulations around advisory. I mean that's the typical European banking uh authority guidelines that we need to follow. So I think there said we built a lot of uh materials around that we also cloud based so again we can, we can leverage what the cloud providers have. But again we also need to see you know how I mean that's typical enterprise SaaS. Of course now with EU AI data act um where you have these you know high risk, low risk, uh classification. Yeah they literally said that loan approvals uh, so fully automated loan approvals using AI are high risk and need to go through a conformity assessment. So now we are really thinking about how do we make sure that we do not get into that category Because a conformity assessment is going to take two to three years to get approved and banks are not willing to implement that. So I feel that it's like becoming a separate capability within our company to bring that assurance to clients, to bring that assurance to banks to make sure like look, we figured this out, we're doing this fully in line with regulators because otherwise uh, they will never implement our technologies. So that's I think a key challenge that we see today.

Speaker A: And then you create kind of because if uh, it gets the high risk stamp, banks you mentioned bank do not want to, are not really eager to integrate it.

Speaker B: Well the thing is then, then you get a bit into the, to the innovation uh trap right? You come to a bank, you say hey look at this, this can save you x amount of uh, of costs or you can be much more efficient like this. And they're like by the way we need to go through a conformity assessment. It's going to take three years. The process is not well described. Yeah, then I think then I've been in this industry now for 15 years then I know what's going to happen. They're going to be like it's very nice, see you in uh five years. So for us it's really about I uh think we're very focused on de risking and that's also what we did with our fully roadmap. So we've pre vetted this with specialists like regulatory experts, lawyers on top of that. So that today the functionalities that we have in the product we can basically go to the compliance department of a lender of a bank and just say look this and this and this and this all ticked off. This is approved in line with regulations, not high risk. Human is in the loop, et cetera, et cetera so that we can bring that innovation quicker to them. So it's actually we productize it in

Speaker A: a sense our audience also can more relate to payments and fintech. Is there more in the future, more collaboration or synergy between mortgages and payment solutions you think?

Speaker B: Yeah, I think for me it's um, I mean a third trend but it has started and for me it just needs to start maturing is data sources and for me payment providers are still really really good providers of uh, data which is very valuable into a loan application etc. I think payment providers today know if people pay back things regularly, um Also payout schemes by the way, for mortgages in, when you go for a renovation project is something that is today also. It's a medieval process actually.

Speaker A: And how do you get them to help or facilitate opening up that data for companies like yourself or the banks?

Speaker B: How do we get them there? I think it's about, uh, that's really about synergies and collaboration. I think some of them have already very good established relationships in banks. But then I think it is more about showing those use cases to the banks and they're mostly sitting on clean APIs. So actually very quickly you can show something that's extremely valuable.

Speaker A: So uh, there's a lot to be won there.

Speaker B: Yeah, I think on a macro level I think we will be, I mean we've seen interest rates went up to fight inflation. The expectation was to cool off housing markets, um, let's say in the areas that people really want to live, like where we're sitting now today, that doesn't happen. So for me there's actually a huge risk on affordability. But the problem is affordability is going to be more hard, but it's going to be more tough to, for, for borrowers to borrow and banks are going to find solutions for that and maybe the government will come with subsidy schemes, etc. So for me what, what that will lead to is actually that we will not be able to. It will, it will prevent us from going to full online mortgages. You just do it very quickly on your phone. It will actually lead to you talking to a broker or a loan advisor much more quickly because we can't figure out how we will, how we will structure more complex products, difficult affordability. So I think I actually see a renaissance moment for the loan advisor. All the banks have been shutting down branches, so your typical around the corner loan advisor doesn't sit there anymore. Uh, but we're going to have to find digital solutions, going to find very nice ways of interacting with humans and basically getting through that process. I think that's where I see a huge area. I uh, mean my prediction is that um, the banks have figured out tooling for that virtual advice, partially digital advice. I think they'll win in the market. And then we also have the whole green financing, renovation, et cetera part which I think will also lead to that. And what I hope is that we will also start seeing the first real embedded experience because I do feel, um, there is things that are going to happen to really have an integrated property buying experience, um, which again in a more complex housing market might Make a lot of sense as well.

Speaker A: Thank you. I think it's good to have a few takeaways of this uh, discovery reading. Ah, the different data sources and making decisions based on that can really impact the business significantly. However, the road towards that is quite long. Lengthy models need to be trained, regulations need to be overcome or are becoming more challenging. Actually the payment industry and payment service providers can have provide a significant impact by providing data source, clean data sources. But in order to really create a more, as you call it, embedded experiences for consumers, there are multiple industries that need to collaborate like the proptech industry. And then, and so far I think the biggest game changer has been open banking which made it possible to at least digitalize some of the mortgage industry. Where we are today is still that there are too much providers, too much technologies that are not connecting well with each other by only using the potential 5%.

Speaker B: I think that's uh, that's, that's fully correct. And I think also the mortgage industry can learn from the payment industry. Um, but it's just. Mortgage industry is always the one that comes last. I have the feeling when it comes to digitization it's also a. Payment happens almost instant and a mortgage takes 20 years.

Speaker A: If I talk with people on mortgages, people look a bit better. Look, really look at this. It's painful. What do you like so much about this mortgage industry?

Speaker B: I think I became a domain. I mean a. I learned uh, although I learned uh, that Warren Buffett doesn't agree with me. So I don't know if you want to go for believability if you need to believe me. But I think I rolled into it because I was a domain expert, let's be honest. And I hated my own experience and I felt we could do it better. But secondly is that something I learned over time is actually that a mortgage is, actually has some societal value because actually people, the first thing they always pay back is their mortgage. Hopefully property prices rise. So in the end you're creating wealth so you're saving money. Uh, unless you're in the Netherlands with the interest only loans, but anywhere else you're paying back for your loans. So you're creating home equity and this can be your retirement. And I think that is something that has a lot of societal value. And actually should I say working also with more regional banks that really have a regional responsibility. I started really thinking this is actually a very meaningful problem, a very meaningful product for households to create household wealth. Uh, it's not only about having a roof above your head. But it's also about, in the end of the day, having something saved up. Which you don't have if you rent the rest of your life.

Speaker A: Correct. Thank you.

Speaker B: My pleasure. Thank you for having me. Voices in payments Expert insights for a digital payments world.