Ritesh Singhania, Co-founder and CEO at Zango: Building AI agents for compliance

0:00Ritesh Singhania:Hello and welcome back to the Scaling Europe show. I'm Seb Johnson. I'm here with Ritesh. Ritesh, thank you very much for joining me. How are you doing today? It's a pleasure. Thank you so much for having me on the show. It is my pleasure. Now look, you are building Zango AI, which is like an AI native compliance infrastructure. Can you touch a bit on what that means? Yeah, so at Zango, we are building AI agents for risk and compliance teams in financial services. So, I mean, compliance can mean a lot of different things to different people. we've had the Want Us and the Trout Us of the World in the security compliance side, and we've had the OnFeedos and a lot of the solutions on the KYC and AML side.

0:37But there's one part of compliance which has sort of always been people-driven, so whether it's your external consultants or your internal teams, and that's regulatory compliance. So when the regulator is putting something out there, what does it mean for the business? Does it apply to them? What are the new rules that they have to follow? and based on the new rules, what are the gaps in their current policies, in their controls? So that part of compliance has sort of always been consultant-driven, people-driven and that's the piece where we are really solving the problem for the financial services.

1:10Ritesh Singhania:Why did you choose to solve this problem? What about you and your background made this problem the problem that you wanted to solve? Absolutely, yeah. So for the last 10 years, I've been building in the space of compliance and what I've always seen is that you've had different point solutions that can solve part of the problem. But for you to be able to go to a client or to a bank or an asset manager and say, this is what has changed and this is what you have to do to make sure you can prove compliance in front of the regulator, software was just not able to do. We cannot get software to read and interpret regulation and be able to apply it to the risk framework of a bank and say, look, this is what has changed and this is what you need to do to prove compliance.

1:53So that is something which just software was unable to do. You could just do point solutions. And with LLMs and the semantic search that has improved, which means that they can understand the intent behind why the regulator is saying something. So those were some of the nuances which you are able to do in a software and with humans in the loop. You can go and give a lot of comfort to the clients to say, this is what it means for you.

2:18Ritesh Singhania:Got it. That's super interesting. And you've had a successful year to date. You know, you raised what, like a$4.8 million round in July. Can you touch a bit on that round? You know, it was$4.8 million. Who did you raise it from? How have you been spending that money? Absolutely. So we raised it mostly from US investors. So South Park Commons, Nexus Venture Partners, they were the sort of the funds we raised the money from. In the UK, we had no label ventures. Then you've got really good angels as well. We've got Alan Morgan from MMC Ventures. We've got Richard Davies, who's the CEO of Alika Bank.

2:55And we've got Notion as well. So it's a really good mix of funds. If you think about it in a compliance space, it's the people who really pay for compliance are the financial institutions. It's not the smaller fintechs. So it's a lot about the biggest risk to the business as you're launching this business is that can you actually win the trust of a mid-market or a bigger bank? to say, look, they're trusting Zango with their compliance. And once you build that trust, then it's a lot easier to go and demonstrate that to the other clients because compliance is a very risk-averse space. No one wants to be the first mover advantage.

3:34So for us, what really worked was in the very early days, some of our early customers were some of the largest banks in Europe. And the fact that they were already trusting us with their compliance meant that it was much easier for the investors to underwrite that risk.

3:50Ritesh Singhania:God, that's really interesting. And I thought there's a couple of things I want to get into on your round specifically, but you mentioned that most of the investors are US. Why do you think that was? From our perspective, I mean, there are a few things I feel like, you know, when you, people always say that you have to go to the US to sort of, that's where the market is, that's where people pay for it. But I feel like compliance is one of the few spaces where you might be better off starting in Europe slash UK rather than the US. Just people's appetite to pay for compliance and how much they think about it as a topic as compared to the US, I feel there is more awareness here and people's appetite to solve that problem and pay for it is a lot more.

4:35So from a US investor's perspective, it did not really matter whether we are UK slash Europe or US because there's an equally or if not bigger opportunity in Europe for people to pay for it. If you think about it from the other perspective as well, people say in Europe, the hardest thing is that because every geography is different, every country is different, and you have to have a different playbook as you're going from Germany to France to Netherlands and so forth. It actually also works in our favor because when you are a fintech and you're expanding to 20 countries in Europe, you need a solution to help you to expand these countries faster.

5:09So for a lot of these reasons, us being in Europe slash UK made it more interesting for the investors, whether they are US or Asia.

5:17Ritesh Singhania:Interesting. And you mentioned some of those really high profile angel investors that you got on board, you know, like Richard Davis of Attica Bank. How did you go about securing and sourcing those types of investors? It's a combination of three things is how I think about it. One is who is making the introduction? I think that matters so much, which is that, look, I have already invested in Zango and Richard or Alan, would you like to have a look into that? So that sort of lays the foundation for that. And sort of the other two things is that have they as founders or investors, have they lived that problem in their life?

5:57So Richard runs a bank. Alan Morgan used to be the senior partner at McKinsey for a long time. And all of these challenges, he's manually sold consulting around all of these all his life. He's invested into a compliance business 20 years ago. So a lot of these things means they've lived that problem and they really want to see someone solve this problem. So it's more of that personal mission, if you like. And the third thing is, do I back Zango as the team to solve that problem? So I feel like it's a combination of those three things. And if you get all these three boxes staked, then it's a lot easier for someone to back you.

6:33Ritesh Singhania:And the way that you get in there initially, is that, like you said, through, you know, through an angel investor, you've already got on board. And they say, well, look, I know, Richard, you know, I'll introduce you. Yeah, absolutely, man. Like, it's a lot about, that's how you build credibility. Because no, like we're an early stage business, very new, haven't got as many proof points as a lot of people would like to see. and how do you how do you risk it for them it's it's those introductions it's someone they already trust and if they are trusting you then it's a lot easier for them to trust you as well whereas i think it can really work the other way around in the way that someone who has decided to not invest in you and they make the introduction i think then you lose all credibility yeah that's yeah that's a great point and how did you find the process fundraising overall was it something that you found particularly tough i mean it's interesting because you're sort of building you're building an AI tool in fintech you know London had this amazing fintech moment it still is having it to an extent but now AI is all the hype you know were you able to leverage best of both that you are sort of you're working with a lot of the top tier banks so you're sort of fintech focused while being AI you know how was the process overall how easy how hard was it fundraising is is a lot about how you actually run the process in the way that I believe that fundraising should not be a super long process.

7:53It has to be run as a super tight two to three week process where you're lining up the meetings with the right investors, with the right people in those funds. And it's a lot about figuring that answer to that question out in the first week, which is that, is your story landing? Have you got the right meetings lined up or not? And if those two things are sort of not haven't clicked, then I don't think a founders should go and fundraise because otherwise you're just dragging the process for six months and the honest thing is that if you're doing it's a full-time job if you're raising money for six months that means the rest of the business is suffering and if the ceo is responsible for fundraising and sales that means you're not winning any new clients so so what for us what really worked was south park common our first investors they were one of our biggest champions they are the ones who helped me some of the introductions to the funds follow on and nexus when they came in, they wrapped up the process within a week.

8:47So from the first conversation to the term sheet was a week-long process. So it's a lot about your story is landing, you've got the right people around the room, and then it's just a matter of just finding the right person or the right fund.

9:01Ritesh Singhania:Interesting. And then what kind of value do you expect? You know, the different investors that you have on your cap table to provide, you know, you mentioned some of the industry veterans are really, you know, they know this problem in and out how involved are they going to be versus the the bigger South Park Commons you know US focused you know what kind of value do you see from from the the range of investors that you've got on your cap table yeah the two things there the first one is that you have to know in your head why you are bringing a particular fund to the round is it because of the strategic value they provide is it because you really like the partners you want to work with is it because it's a fund which can write follow-on checks.

9:41I mean, expecting someone who is a massive US fund who might not have as deep connections in Europe, expecting them to open a lot of the doors is unrealistic. So I think it's just being very clear in your mind as a founder to say why you're taking money from this particular fund or individual is extremely important. But as with some of the strategic angels, it's either because they understand this space really well and can make some really strong strategic introductions. Or in terms of how do you sell to this client base? They've been on that journey before and can help you with a lot of the GTM questions and distribution.

10:16Because at the end of the day, the biggest problem here is around distribution. How do you win the trust of a large bank? And what are the sequence? And how do you think about building the sales team, the distribution channels, the GTM? It's what some of the people help you with. Like a lot of the conversations with Alan Morgan is about understanding how do you win the trust of these clients? What kind of segments do you put them on? Because that's what he used to do for a long time with McKinsey. That's sort of the last point around the investors and the value is at the end of the day, they have written 20 to 30 to 40 checks to different founders.

10:52It's a lot about the relationship that you build with them. How much of a champion they are of Zango as compared to another business that they've invested in. And so a lot of it boils down to the relationship that you can build with them. They're like, look, I would really love for Zhang Wei to succeed. I'm going to see how else can I help them. So it's a lot on you as a founder or the founding team to build that relationship with the investors as well, where everyone's asking for their attention. Where do they decide to put their attention?

11:21Ritesh Singhania:And you kind of mentioned there how important it is to get introduced and to sell, I guess like selling to and learning about how the big banks operate and how procurement operates, I guess. How have you found it selling a tool like this to, you know, very large, often, yeah, you know, stuck back institutions that are often stuck in their ways? You know, what's it like trying to send innovative AI product to quite an old industry? I think right now with the combination of two things, Seb, one is that people have been oversold AI, especially in the legal and compliance space where the AI still hallucinates.

12:01that still make mistakes. And going in with a pitch that, look, I'm going to automate your compliance just doesn't land well. So one is that managing their expectation around what you can do and how you can help. And the second is they don't know you at all. Like everyone wants to do something in the space of AI, whether it's your compliance team, risk teams or legal teams. But why should they trust you? I mean, if something goes wrong, they're going to lose their job.

12:27Ritesh Singhania:Yes. Risk and compliance, it's not just anything else. So how do you build that trust? How do you become that trusted partner? And I feel that the first meeting cannot be a pitching the product meeting. The first meeting has to be building that trust, building that credibility, building that relationship and helping them understand who you are and how you work. And then over a period of time, when they're ready to do something in the space of AI is when they would come to you to say, look, Zango, we've got this problem. Can you come and talk to us how this would happen? so it's about making sure that you build enough trust and credibility that when someone is ready to do something in this space they come to you first because they trust you more than someone else it's like the consulting space like you go to your consultants when you have a problem because you trust them and how do you become the trusted partner is is a lot of what the journey for us is about and how do you balance that with the need of the need for super revenue growth that we're at the moment you know everybody's trying to get to 5x ARR in 12 months and crazy milestones whatever and you're trying to say you're saying that you're trying to build long-term relationships so that when you need they need you you're there yeah so for us it's a slightly different business in the way that uh we we might not be able to win clients at the same pace at some of the other startups but the two things that work in our favor is that once we win the trust the contracts can go from 50k to half a million within 12 to 18 months.

13:57Wow. So it's the land and expand motion, which really works. And the second is when a financial services or a mid-market bank works with you to integrate your mission into their current stack, you're there for life.

14:11Ritesh Singhania:So very sticky, very embedded, yeah. Because I'll give you an example. With one of the clients, it took us 11 months to go from proof of concept to implementation to complete success of the solution. Out of that 11 months, it took us six months to convince their IT team to open their APIs, trust how we are doing the solution. So because for the business teams, the amount of time it takes to get a vendor in the procurement process and integrate with their current stack, the next time when they have a problem, they'll go to the current vendor rather than trying to find another solution. So even if another solution is 2x or 3x, better than yours it's just not worth what they're trying to go to another provider so just very very sticky got it okay so it's more of a classic sas enterprise uh relationship where you know you you yeah you invest in doing the deal getting it done land and expand and you know that you'll stay there for a long time yeah um i also want to talk about the tech that you're that you're using itself so i've read that you're kind of using sort of your own llms to do a lot of that work can we touch a bit on you know how proprietary are your models versus sort of general purpose ones yeah yeah so we're using the foundation models we're not uh creating our own uh foundation models or llms the way where we are different than any of the horizontal tools always that exist out there is when a regulation comes out how do you read and interpret that regulation is very subjective it's never white and black.

15:45So if you're asking the Gemini, HRGBT, or Covara to sort of interpret that regulation, it does sort of its own thing, and you can never rely on that. Whereas, I'll give you an example. For example, when FCA says something, it's very outcome-based. It doesn't tell you what you need to do. It tells you what outcome you have to achieve. But in order for you to understand what you need to do, you might have to refer to five different documents. You might have to refer to the FCA handbook. And that's different for FCA, that's different for EBA, and that's different for SEC. So how do you build in that intelligence into your model that when something comes out, how do you actually read it?

16:22And what are the other documents you might have to refer to be able to go to a client and say, because of this, this is what you have to do. So that's the first thing, which is where we build our IP. The second is, every bank is different, their business model is different, their risk appetite is different, their products are different. How do you personalize it for them? How do you train the models on Monzo's business to Revolut's business to Starling Banks to HSBCs? They all have different risk appetite in how they operate. And how do you build in those nuances into the model to actually, that's what the consultants do.

16:56They tell you, look, it's not what you have to do. It's like, look, for your business, this is what you need to do. And that's where we come in, where we don't just give you like 20, 30 percent efficiencies. in your compliance, but we actually tell you in a very personalized way what you have to do.

17:13Ritesh Singhania:Amazing. Got it. Okay, that makes a lot of sense. And how have you found sort of that use and adoption to be within the teams that you're already live in? You know, how well have teams been able to pick it up and run with it? The thing that makes people very, very nervous is if you go in and say, I'm going to automate your compliance, because what that means is that it means you're taking away the control from them, because at the end of the day, if a compliance officer was to sit in front of the regulator and demonstrate, look, we've got the right controls in place. So the moment you start the conversation on, look, let us come and automate your compliance, it sort of doesn't really work very well.

17:49What really works is that if you can help them with reducing their risk or making the controls stronger, which is that, look, when you sit in front of the regulators, you can actually demonstrate better compliance than what you did back in the day. And what that means is, look, back in the day, the compliance team was only doing sample checking of some of the data points. Now, because of using our solution, you can do 100 % testing. So how do you help them have a conversation with regulator, with the board to say, look, we have got better governance in place, rather than we have automated our compliance?

18:25And so what are those? So I'll give you another example of the use case. If you think about whether debt collection, onboarding, credit underwriting, a lot of these processes will be automated by agents in the back office. Can you build an AI governance layer that sits on top of these agents, which are continuously monitoring the job of these agents to make sure they don't go rogue? So how do you build it? That's what I mean by compliance infrastructure. You've got agents who want to call the people who have defaulted on a loan and say where things are. How do you make sure they're not saying things that they shouldn't be saying, which ends up in a lawsuit.

19:01So it's about building that governance, more control, which gives compliance teams more comfort. And that's when you win their trust to use them. But otherwise, it's a slow moving space. But the thing sort of that helps is that it's a time where everyone's curious. They know the nature of their roles are changing. They know how they are doing things is changing. It's just that revient on their own journey of adoption.

19:27Ritesh Singhania:Yeah, and I think we're in a very different place in the market than we were a year or two ago, where now everybody in their various jobs is starting to think, okay, I need to be thinking about AI in some way or another. And if I'm not adopting it soon, I'm going to be falling behind, which I think is really interesting. And I think it's creating a lot more demand for AI startups and their products, which I think is a great thing. I want to ask that one final question, which is over the next 12, 18 months, or as you start to look forward to that next round of fundraising. What is it that you need to achieve to make it a success?

19:59Yeah, it goes back to the two things that I was saying. One is that, can we demonstrate that land and expand motion? Which what it means is that are we actually able to build the trust of the clients that we are working with, where they come back and say, can you solve more of our problems? So which means that we become the trusted partner to them, just like how the big four is to the financial services right now. So can you start to become that trusted partner to your existing clients with the trust you've with more problems? And the second is, how do you go from mid-market to enterprise, the bigger sort of tier ones that are out there?

20:38And are you able to win their trust as well? Because ultimately that's what matters. Like there's something which someone said to me, they were pitching to a tier one bank and they lost the contract. And they said the reason why they lost the contract because the CEO of the bank thought that the regulator doesn't trust them as much as they trusted the competitors. Wow. So it's a lot about, like everyone's got skeletons in their closet. It's just about how deep do I want to look? And because I'm using, let's say, Zango, I already trust you.

21:08Ritesh Singhania:That's so interesting. It comes down to really relationships and reputation. And if you haven't built relationships, if you don't have a squeaky clean reputation, it's going to come back to haunt you. well look thank you so much for taking time to chat ritesh i love chatting it's very cool what you're building so keep me in the loop of developments or changes uh yeah i'd love to i'd love to cover it cover it in more depth absolutely man and thanks a lot for the questions no my pleasure all right let's stay in touch