The following interview is a conversation we had with Fabrice Deprez, CEO of Discai, on our podcast Category Visionaries. You can view the full episode here: $12M Raised to Build the Future of Anti-Money Laundering Software
Fabrice Deprez
Thank you, brother, very well. And how are you?
Brett
I’m excellent, thanks for asking. And super excited to chat with you. So am I. So let’s jump right in. So, to kick things off, could you just start with a quick summary of who you are and a bit more about your background?
Fabrice Deprez
Okay, well, first of all, like you said, my name is Fabrice Deprez. I’m approaching 50 years, still in the 40 train, like we say here in Belgium. I’m original from a family of entrepreneurs. So my father had a company, my grandfather had a company, and it has always been my ambition to manage successfully another business. And for that, I start my career in consulting. So I’ve worked around 20 years in consulting business, big American companies accenture, CNC, Deloitte, and so on. But I would say six years ago, I had the opportunity to manage a small It company which was, I would say, in financial difficulties. And we brought that company back with the team over a period of five years, back in business with a lot of growth potential. And then a little bit more than one year ago, I was approached by the KBC Group, So, a large financial institution in Europe, which had the idea of putting this kind of market company.
Fabrice Deprez
Amesh and I launched the company in March 2022. And here we are.
Brett
Nice. That’s amazing. Now, one thing we like to ask, or a few things we like to ask just to better understand what makes you tick as a CEO and as a leader. First one is what CEO do you admire the most and what do you admire about them?
Fabrice Deprez
Well, the funny part is that the CEO I admire the most is basically the CEO of KBC. I already had it in mind for 20 years. I had the chance, as Accenture consultant, to do a big project in the group in the early days where Johann was already working. And I was basically impressed by his speed of reflection, the quality of the discussion, the impact, but also his way of approaching the things. So I would say almost 20 years later, when you get a phone call to have a discussion with him for a potential job while you just jump into the opportunity. Johan is managing a company of 43,000 employees, one of the most valuable financial institutions in Europe. So when you had a chance to have a chitchat with this kind of person, you just jumped into it. And at the first interview, first discussion, he reblowed my mind about the way he approached, the way he discussed, the way he motivates, but also how he’s able to combine vision, which people insights, motivation, drive.
Fabrice Deprez
It’s impressive.
Brett
That’s amazing. And can you give us more context on the parent company, KBC Bank? How big is this? Because a lot of our listeners are from the know, we’re going to know Chase and Wells Fargo and I guess Silicon Valley Bank now. But can you give us an idea of the size of KBC Bank?
Fabrice Deprez
So, KBC, first of all, originally it’s a pure Belgium bread bank, but we have also a special model. We are bank insurers. So we have a bank branch both on retail, corporate, private banking, and we have also insurance entities. The company has 43,000 people sorry, spread over six different countries in Europe with a big focus on innovation, digitalization, and drive of customer attention through new technologies.
Brett
Fascinating. And I know at the start there were talking because of the structure here, you’re not technically the Founder of the company, you’re the CEO of the company. But I have to ask, do you still consider yourself an entrepreneur? Because this feels like an entrepreneurial endeavor, right?
Fabrice Deprez
Yeah, absolutely. Because the story behind the company is that KBC, like I said, wants to be innovative. So we are investing already for 15 years in cloud, ten years in AI, artificial intelligence, machine learning, and we have developed many applications. At one moment, we just recognized, or they recognized, because I was not part of the ship at that moment, but they recognized that a number of these applications were quite unique on the market. It was confirmed by many strategy companies saying that they didn’t see this on the market yet. And of course, that raised the idea, while we can market that and generate a new revenue stream. And this is how they came to me and say, well, we have that idea. The company is in the first step of creation. Do you want to get that company on the feet? So we had to build an organization, we had to attract people.
Fabrice Deprez
We had to develop a commercial plan, financial plan, and just go on the market and put the name on the street. Like I said originally, we are in the regulatory compliance part, which is quite securized, which is quite, like I said, regulated. So as a new name, with the new people, with a new product going along in this kind of very stable and controlled environment, well, yeah, I can say that we are still entrepreneurial. We are still building step by step the company.
Brett
And another thing we like to ask, just to better understand a bit more about you, is about books that have had a major impact on you. So this can be classic business books or it could just be a personal book that you read that really influenced how you shaped the world. So do you have any books like that come to mind?
Fabrice Deprez
Yeah, that’s a very American book. It’s very American, very famous in many aspects from good to great. I think that is one of the books that is near the bed of many consultants and people in America willing to do management at a good level. I read it several times and there are two elements that stroke my mind I still keep in my management is that in the book? And they describe how it’s more important to have the right people and based on the right people, design or define what you’re going to do with them than defining a job and then looking for the right person. That is one thing what I’ve always used in my career and which makes really sense and can help us to make a difference. So that is one of the aspects. And then the second aspect which I think that many managers forget is that you are replaceable dispensable.
Fabrice Deprez
So it’s always the idea that you are always thinking what would do the company without me and how they would grow without me. And these are two things where makes I think the difference. And also in our little team, it’s always an aspect to have the right people, not the right job and thinking that everybody is disposable. And if one of us would disappear the day after, the week after, month after the company need to be stable and continue on.
Brett
Nice. Love that. Good classic American book, like you said.
Fabrice Deprez
Absolutely. Yeah, absolutely. But it’s really aspiring. Even it’s more than 20 years old, it’s still relevant.
Brett
That’s what always blows my mind a little bit. I’ll read some of these books that are like 20, 30, 40 years old and somehow they’re still relevant and it’s really timeless principles that haven’t changed, which I think is just fascinating to think about. When you really boil business down, it’s the same stuff no matter what the industry is, you can really boil it down to the basics.
Fabrice Deprez
Yeah, well, because in any business, in every business, of course the product is important, the market is important, but in the end, everything what it turns around is around people. Even in this digital area, it’s the people that counts. And I’m still convinced and I’ve seen it through my career, if you just bring the right people together, you can achieve anything.
Brett
Yeah, amazing. Now let’s switch gears and let’s talk about the company a bit more and more specifically the products. And it sounds like from the start of our conversation that there may be more products in the works, but I believe today it’s the AML solution that you have on the market, is that correct?
Fabrice Deprez
Yeah, absolutely. So to make a long story short, because I could of course speak for hours but it won’t be a secret, even for people not in the financial industry, is that compliance is getting more and more important. The regulations are a burden one side, but on the other side it’s a saving. We’ve seen it in Europe, but what have happened last few days with the banks in the US and then moving to Switzerland and getting closer. But the regulations that have been put in place have supported heavily into the protection of the bank. So that is one really important thing. And most of the people think regulation about financial stability, liquidity of a bank, that’s one side. But of course on the other side there is also everything within the financial crime. So the financial crime can be put into two areas is first, protect the bank and the customers against fraud or any misconduct or protect the government against thefts like money laundering is doing.
Fabrice Deprez
And this issue is, I would say, underestimated. It’s a huge problem on the financial market. It’s a huge problem for countries to get the right tax entrances and regulators have been pushing a lot of this work to the banks. So banks are obliged to analyze a lot of what is happening through their books in order to identify potential money laundering. So to give you a rough idea, we have seen that in countries like the Netherlands, like Belgium, it’s multi billions of work that is done by the bank in order to comply with regulation about identifying money laundering patterns. Right. And money laundering, for the people that don’t know it’s, just trying to get money from A to B without paying taxes. And this generates, this process of analyzing is partially automated but also still a heavy manual work done by investigation teams and this creates a huge workload on the bank.
Fabrice Deprez
So current systems are, I would say, not effective and efficient as they should be. Efficient means it generates a lot of manual work. Effectiveness means that able to identify these money laundering cases in the books and current systems are just not relevant anymore. What we have brought is a solution which is AI based, which is I would say mimicking and automating a part of the manual work which is done by the people simplifying it, automating this part, which is drastically reducing the manual work. So working on the efficiency but also supporting, thanks to technology identified patterns which were not detected until now. So working in both aspects, efficiency and effectiveness.
Brett
And if we’re looking at the banks, they’re under a lot of pressure right now. Right? And the stakes are very high. I was reading a report in the media the other day and it was, I think like two or $3 billion in fines last year alone, which is obviously bad. And I think I read that Credit Suisse was one of them. Right. That was one of the reasons that I think partially led to their downfall in the recent years and put them in this vulnerable position was they had scandal after scandal. There were a lot of fines. This is high stakes stuff for these banks, right?
Fabrice Deprez
Absolutely. Because in the end, what can make a bank fall is a run of the bank, meaning that people don’t trust the bank anymore, they deposit back and then the company has liquidity issues and they can’t fall. So of course, the trust of the people in the bank is really important and trust is linked to many aspects. So even bank at one point gets into the newspaper with fines, in money laundering, in fraud or anything. This just attacks their credibility, their reliability, and therefore could result at one moment into a fear and therefore a run of the mac.
Brett
Yes, makes a lot of sense. This show is brought to you by Front Lines Media, a podcast production studio that helps B2B founders launch, manage and grow their own podcast. Now, if you’re a Founder, you may be thinking I don’t have time to host a podcast, I’ve got a company to build. Well, that’s exactly what we built our service to do. You show up and host and we handle literally everything else. To set up a call to discuss launching your own podcast, visit frontlines.io podcast. Now back today’s episode. Now talk to us about adoption as you’ve brought this technology to more banks, what’s that been like so far and are there any numbers that you can share that highlight some of that growth?
Fabrice Deprez
Well, let’s say that in general there are some figures where are generally accepted in the financial industry that I would say on effectiveness is when you look let’s go back in the process for people that don’t know what is the AML anti money laundering detecting process? First of all, the bank need to analyze their transaction. They need filter transactions and identify potential fraud cases or money laundering cases. These lists of transactions are then investigated manually by investigators and they have a certain ratio of in all the transactions that they receive. What is the ratio of false positive? True positives? True positive is it’s an alert and it’s indeed a case that needs to be reported false positive. It is an alert, but there is nothing at the hands. So this ratio we see at many banks that they are struggling with 5-2-7 of true positives.
Fabrice Deprez
So more than 90% of the work that the people are doing is just a false positive, meaning a work which was unnecessary. When you relate that with what is the percentage of what can be identified of money laundering is banks are struggling with 1015 20% of detected cases. So if you put that 90% of the work is non relevant and whether true positive they identify, they just are able to identify 1015 20% of the true positives meaning money launder has between 510 and 20% chance of being caught when they launder their money in a book. So these ratio are really important with using AI in a good and decent way. We are with our solution, we are able to almost multiply this factor by three four or have in total three multiplied by three better performance of up to 10%, ten times. So we could reduce drastically the matter of work or increase drastically the number of cases identified of a mix between that and that’s.
Fabrice Deprez
When you look at banks on US level or large bank and European level, this means saving billions on the country level. So, yes, that’s quite an important in Belgium, we’d estimate that only 5% to 10% of money laundering is detected and is collected back on tax level. So if you take a country which has a turnover or GDP of 600, 700 billion and only 5% to 10% is identified, and you could go to 50% of identification and you can make accounts of what tax level could be recuperated on the country level or the savings that could be done by a bank.
Brett
Fascinating. And another question, because you’re technically part of a bank, are your customers then competitors of the bank? And if so, does that create a strange dynamic in any way?
Fabrice Deprez
Yeah, that’s a good question, Brett. Well, when we launched a company, we had the choice to launch a company with different domains. All process, efficiency, commercial activities could be improved thanks to AI. But we absolutely made the choice to launch the financial crime roadmap and suits of products because we said just as a bank, well, this fight against the financial crime is a common fight between all banks. And this doesn’t bring any commercial advantage. Of course the bank is making savings, but on the other hand, it’s a common obligation to do that correctly, just to, I would say, keep the trust on the market. That the banks together are trying to solve that issue of fraud, money laundering just for reputational damage. It’s never interesting for a bank to have a big hit on the reputation of a bank near to you because it’s like the domino effect, that reputation won’t reflect on you.
Fabrice Deprez
So this fight is common. And in that optic, then the bank are all the customers we visit. So we started one year ago. We visited already 26 banks spread over Europe will move to the US and Canada after the summer. Every time we receive that same answer, yes, we imply that it’s a common fight and we should fight that together.
Brett
Got it. That makes a lot of sense.
Fabrice Deprez
Yeah.
Brett
And I think we’re seeing that today, right? How interconnected banks really are. If one bank has issues, then every bank has issues. At least that’s what it seems like in recent weeks.
Fabrice Deprez
Yeah, absolutely. Like I said, it’s a domino effect and everything is a question of trust. So it’s easy when one bank has a problem that you come in the newspaper saying well, the other banks in the same country, well if one is bad, well, they should be all bad. Which is I would say not always true, but in the end it’s how the public reacts because this is how the public reacts can generate at one of the banks. So it’s all in our common interest to keep the trust as high as possible for all the banks, at least at this level.
Brett
And as you move country to country so you just mentioned you’ll be moving into the US. There, as you do that, does the algorithm have to change? Is it different or is the AI different when you move to the US from maybe Germany for example, or is it all the same thing and no adaptations need to be made at a country level?
Fabrice Deprez
Again, a good question. Well, in the AI there are two things is the model that we develop, which the code that we develop which generates and the models on a bank. Now, the basic idea behind a solution is that we do customer profiling in a very detailed way and experienced way with all the data available in the bank in order to define suspicious behavior and therefore suspicious people. Now that process of the code in order to identify is the same. But of course when we train a model, the model will be trained bang by bank because of course the topology of a customer of a Wells Fargo won’t be the same as KBC. I would say in Belgium the topology of the people of the customer will be different. On the other hand, the solution just adapt itself to the topology of the customer and therefore into the process.
Fabrice Deprez
So the code that we develop is the same but the model that we will be training will be banking specific.
Brett
Got it makes a lot of sense and whenever you talk about AI, I feel like it has to be brought up of potential displacement of workers. So how do you view that here at some point? Does this end up automating and taking away some of those jobs of KYC and AML professionals? Or is this something that more enhances them, to free them from low value manual work and focus on more important work?
Fabrice Deprez
Well, it depends a little bit how you apply it, but I’m convinced that in the end it will generate a lot of work. Now, this is a comment that I read a lot into articles and so on and indeed, of course the program need to be programmed, the models need to be retrained, data needs to be accessible, but this is quite, I would say, putting work on data, science on it work and so on. This ID doesn’t contribute into the other types of work. Now it will do it one side. On the other hand, what I’m also convinced is that twofold. First, all companies, all organizations are struggling to identify and to find the right people that are willing to work on certain subjects. So there is a lack of resources and we can cope with that a little bit better using international technology for that.
Fabrice Deprez
And the last thing is, what we have seen a lot in Belgium is that there has been or in Europe there has been a flood of resources or move of resource to low cost countries. So, of course, a lot to India, a lot to China. And when you move that, you move all the layers from the executional work, to the semi intelligent work, to the high intelligent work, and you move everything at once because otherwise it doesn’t work. Just moving the low cost, low hanging fruits to an offshore country. Well, in the end, the quality just drops. Now, by using AI, we can automate a big part of the low added value work of the people. And therefore, by being able to automate that, well, it makes financially interesting again to bring the mid level and the intelligent level back work. So if we move an outsourcing, we outsource a part of call centers or manual worker process and this work could be partially be automated, the business case becomes again interesting to reinsource a large part of the work, if that makes sense to you.
Brett
Yeah, that certainly does.
Fabrice Deprez
And that I’m convinced that this will bring a lot of work back to our countries, to our entities, and therefore beneficiary for everybody.
Brett
And I don’t know how it is currently in Europe, but in the United States, AI is everywhere. And it’s been like that for the last few years. But obviously it’s know very loud. For the last three or four months since Chat and GPT came out, all everyone can do is talk about AI. I think there’s just a lot of buzz around AI. What I hear a lot know companies that are just saying we have AI, but they don’t really have AI, things like that. So for you, what do you do to really differentiate and make sure that it’s clear to the market that this is real, this is artificial intelligence, it’s powerful artificial intelligence and it can be a real game changer for the banks that embrace this technology.
Fabrice Deprez
Yeah, absolutely. Well, in order to answer that question, well, we need to understand what we mean with AI, because also in AI there are many flavors, technologies and of course, when you see what Chat GBT is doing, it’s impressive for, I would say, what we first see. Now when I discuss with my data scientists, my AI specialists, they say, well, what they did is brilliant, but in the end, it’s not complex. They have just used a huge load of manual work. They have used a lot of computer capacity and coding. It’s not that it was so easy, but in the end it’s a lot of work to come to that result. So yes, it’s interesting. It’s a game changer, probably. On the other hand, let’s just demystify a little bit AI. AI in another way is a lot of statistics is use computer power to do a lot of correlations, correlation between information in order to take conclusions.
Fabrice Deprez
So I would say AI is less complex, less advanced than a lot of people think. By the way, the AI principle already exists since the years 50 60, these ideas were already the only thing we did have the compute power to manage that. So on that story I would say yes, I 100% comply. You need to say that you do AI to still be valuable on the market. We see a lot of companies pretending, we see a lot of companies testing. The big difference is how far are you in your testing and putting that into production, there is already one big difference. Secondly, it’s how advanced are you in testing? Just to give you an idea, what have we been doing that I think it’s different? Firstly, on the program that we are bringing on the market, we are already developing for five years. So the first three years it was purely development and then two years of testing, tuning, until the algorithms were delivering the result that we wanted.
Fabrice Deprez
And in the end, what we have done is to resolve that. Let’s come back to the money laundering. Like I said, a process and then investigator received a transaction. They need to save that transaction. Well, is that now a true positive? Is that the money laundering case? Yes or no? What does the person does? It goes into the systems or bank and say I will create myself an image of the person. Who is the person behind that transaction? Is it the individual? Is it the company? Where is it located? Is it stable company? Does this behavioral? Is this transaction logic for this kind of person, this kind of transaction? So he generates himself an image of who is the person behind. And this image is based on his experience. He says when a company of that size, of that topology, in that neighborhood, it’s logic that they do that behavior.
Fabrice Deprez
And this is what our company, our solution is just doing, is building huge networks of statistical logic behind. A company of that type at that area has a high propensity of being a bad person, yes or no? And that is simply it. So it’s just statistical regression on a lot, a lot of data, on a lot of historical data. So when you simplify it like that, AI is not complex. The question is do you provide a solution on the market where you build quite quickly a lot of the statistical analysis just to be able to pretend to have it because you need to have it if you make sense. We’ve seen it with Google, we’ve seen it with microsoft after chat GPT where they need to provide something versus having built these statistical regression analytics over multi years, seeing what is true, yes or no?
Fabrice Deprez
So yes, there are a lot of buzz but I still believe that it will move into that direction but it will take time because quality is just tuning, adapting, optimizing like everything with this mathematics and statistics making sense for you?
Brett
Yes, that does, absolutely. I think I cleared up a lot of the ideas that I have in my head and I think all of our listeners are also trying to navigate this world and really figure out how to communicate AI. A lot of companies and founders listening are obviously trying to use AI, but I think every company right now is struggling with how to communicate and educate the market and talk about what that actually means. So your perspective there and insights there are super valuable and super interesting. Now, last question here for you before we wrap because I just realized we are up on time. So let’s zoom out into the future. So three, five years from today, what does this guy look like?
Fabrice Deprez
Well it’s funny because we just had a board meeting yesterday so I had to present what is the ambition. Now to be clear, the company was launched in March last year. We already have our first three customers outside the group which is already positive. We have an outlook of being financially independent as of mid of next year. So two years after launch being profitable, which would be an amazing outlook. In parallel we are working on two axis. First of all the spread on international level like I said, we are already throughout all Europe visible, present and discussing with a lot of banks. We will move to Canada US as of end of the summer. So I would say the area spread and then it’s the products. So we are working on AML but we have a full roadmap of full financial crime which goes from AML fraud and sanctions, cyber protection and everything.
Fabrice Deprez
So we should have by 26 a full pattern of solutions covering all these subjects and in parallel we are evolving into securities domain and into the insurance domain as well. So by 26 hopefully be relevant and be really present and being some I would say in the market leadership on I would say AML financial crime and starting launching new products into insurance and security domain and that on a worldwide scale. That’s the ambition.
Brett
Amazing. I love it fabrice. We are up on time so I’m going to have to cut it here. I’d love to keep you on and keep asking questions so we’ll have to save that for the part two of the interview that we can hopefully do sometime in the future. Thanks so much for taking the time to share your insights here and really educate myself and our audience. I feel like I’ve learned a lot. About inside money laundering and how you phrase what you’re doing in terms of AI was also incredibly useful. And it’s just really fascinating to hear from a CEO who’s building a company within a big, massive, so very fun conversation. I really enjoyed it and thank you so much for your time.
Fabrice Deprez
Thank you so much, Brett, for the opportunity and looking forward to have Banks contacting us just to chitchat. We are always open to exchange experiences.
Brett
Sounds great. Thanks again for your time. I appreciate it.
Fabrice Deprez
Thank you too. Thank you, Brett.
Brett
This episode of Category Visionaries is brought to you by Front Lines Media, silicon Valley’s leading podcast production studio. If you’re a B, two B Founder looking for help launching and growing your own podcast, visit frontlines.io podcast. And for the latest episode, search for Category Visionaries on your podcast platform of choice. Thanks for listening and we’ll catch you on the next episode.
Twitter
facebook
linkedin