Leveraging technology to tackle ESG risk - Charles Delingpole, Founder of ComplyAdvantage













About this episode


A must-listen episode for anyone who is interested in social entrepreneurship and combining for-profit activities with social good.


ComplyAdvantage is a market leader in the use of technology, machine learning and big data to automate and scale solutions that help organisations and individuals mitigate ESG (environmental, social, governance) risk; in the process helping to tackle money laundering, human trafficking and other crimes. The firm is a commercial venture with a strong social purpose at its core and they have secured just over $100 million in funding from the likes of Goldman Sachs, the Ontario Teachers' Pension Plan and Index Ventures.


We discuss a range of topics, from international sanctions and political exposure to adverse media and reputational risk. Charlie provides insight into the company, their solutions and his own trajectory and personal narrative.


About Charles Delingpole


Charles Delingpole is the Founder & CEO of ComplyAdvantage. He is a technology entrepreneur with over 10 years of executive experience in the financial services space across the US, Europe and Asia. Previously he was the Co-Founder and CEO of MarketFinance, a corporate financial solutions company that raised $59M in funding, and before that he was an investment banker at J.P. Morgan Cazanove. Charles also founded the world’s largest student discussion forum The Student Room Group when he was 16. He holds an MA in Politics from Trinity College, Cambridge and an MSc in Management, Strategy, and Finance from the London School of Economics.


Transcript of this podcast episode featuring Charles Delingpole


Alberto Lidji: Well, why don't we start by finding out a little bit about ComplyAdvantage. What's the organization all about and how is it involved with tackling some of these ESG risks in supply chains?


Charles Delingpole: Great. So ComplyAdvantage is a database of primarily people who are bad, so terrorists, money launderers, criminals, but also we've extended that to people and companies involved in ESG crimes.


So people involved in environmental crime, people involved in human trafficking, violent crime. And so, the reason we wanted to come on today was to talk about the new product that we're launching, which specifically helps address supply chain risk with ESG and other such problems. So what we've done is launch a database of every company in the world, along with every director and shareholder, and then map that to every risk possible with ESG.


So the real challenge companies have is with that long tail of companies. And so we've now raised nearly a hundred million dollars and we have a team of 300, mainly data scientists and data experts who are mapping all these risks. And we think this will be a big step change in the way in which companies can help overcome the problems in their vast supply chains and find any hidden risks that they might not be aware of.


Alberto Lidji: So let me ask you, Charlie. If somebody is doing business internationally and their investors or their business partners are asking them, what are you doing about ESG risk in your supply chain? And how are you being active about it? What are the options available for somebody like that?


Charles Delingpole: So, I faced this problem in the last company that I started, which is market finance. And so we were onboarding companies, small companies, and we had to understand their directors and shareholders without having met them beforehand. And often they'd have parent companies overseas or vast chains of companies. And so what we built is a mechanism whereby for any company in the world, you can simply type in the name of the company, find all the directors and shareholders, but also all the risks that are linked to them and all the parent companies. So you can say, okay we have a supplier, suddenly all of their directors have been involved in a coup or one of them is involved in using children in Africa as child soldiers, or one of them is burning down Malaysian rubber plantations.


So I think the old way of doing that was to kind of try and Google things. So you'd often use keyword based matching. So [as a hypothetical example] might Google... if you have a footballer, you'd say Ronaldo terrorized the defence and shot at goal. Whereas what you should be doing is being worried about Ronaldo led a terrorist attack and shot a policemen.

So what we're doing is we're using machine learning as a natural language processing to automatically ingest 15 different languages and then map the entities. And then also link that in a big graph to the companies they control and this connects billions of data points seamlessly, and then understand precisely if this person or company you want to deal with has any risk.


Alberto Lidji: Because the universe of companies out there is massive, right? I mean, one would think, well, how can somebody possibly put onto a database all of the companies out there or the vast majority of the companies out there and provide information that's going to be tangible and actionable for someone who is looking to mitigate ESG risk in supply chains.

Tell us a little bit about the technology underpinning your tool. How does it change the options that are out there these days for helping companies or helping individuals mitigate ESG risk?


Charles Delingpole: So I think the way in which things were done previously was that you'd have separate databases of companies, or you'd you'd have a database or company registry and those wouldn't be connected to the different risk data datasets. And so what we've done is build the first data set, which combines both of them. So we've gone to every corporate registry in the world. We've mapped every company we've gone to hundreds of thousands of underlying sources as well.


So, that sanctions political exposure. It's adverse media datasets, and we built everything from scratch, going back to the underlying source. And then we basically combined and resolved all the entities into single entities and then mapped the relationship. And then we built on top of that, a rules engine, whereby you can specify which risks you care about.

So we have a 16 category taxonomy of risk which covers things like. Credit defaults or environmental crime or violent crime or being involved in the sanction. So you can choose which crimes and which risks you care about. And so a big challenge, particularly for companies that have vast supply bases all over the world with many subsidiaries connections and beneficial owners is understanding which of those could be a risk.


So I guess what you saw last month was the Deliveroo IPO. Many people were concerned around the ESG risk of workers being involved in sub minimum wages. I think with the wave of scandals you've seen like Greensill, Wirecard, like every week, there's a scandal and that won't change.


And so if you want to manage your ESG risk and stay out of the headlines but also, more importantly, not do damage to the environment or society, then you need to understand the people and companies you're doing business with. And this tool has been like two years in the works and that's why I'm delighted to share it with you today.


Alberto Lidji: Tell us about developing it to like this. So you've been around for a little while. You've been incredibly successful in terms of fundraising. Give us a little bit of a flavor for how the company started, the sort of investments that have been required in order for you to develop this product and get all these big brain people onto the team to start developing and coding this stuff.


Charles Delingpole: So, although it's a podcast, we can now look at each other right now. But you can see the room I'm in now. I started the company on this desk in 2014 and down below in my garage, I had seven developers and then I had two people dealing with clients next door. I basically funded it myself at the start.


And then we raised $7 million Series A in 2016, $30 million in 2018 from Index Ventures and then we raised another 50 from the Ontario Teachers' Pension Plan in July of the pandemic year. And the team is now nearly 300 people. So we have teams in New York, Singapore, Romania and London and yeah, I think in terms of the technology... so when I started the company, most of the industry servicing the sector was done very manually. Whereas I think what you've seen is the evolution of machine learning and advanced technologies used by companies like Google, Amazon, Facebook, Apple so it was really about the practical application of that technology to this space and importantly, doing things by algorithm at scale.


So we wanted to kind of make data hyperscale as in just so impossibly large, that it wouldn't be feasible for it to be done manually and we just constantly improved the systems, the breadth depth, speed, accuracy of the algorithms and the step changes from 11 million bad people in companies to every person and company in the world and every risk spanning fraud, credit identity but, most importantly, ESG risk.


Alberto Lidji: Now with ESG, you know, it's difficult to open a newspaper today without looking at a headline that has ESG involved in it. Was that a consideration for you as well? When you're thinking about the sort of entrepreneurial venture that you wanted to be in. Do you like the fact, or were you focused on the fact, that this sort of technology will actually tackle things like human trafficking? It will tackle many other aspects of society that are not desirable at all... is this a driver for you?


Charles Delingpole: So I think the first company I started was about broadening access to education for people who weren't from wealthy backgrounds to go to pieces like Oxford, Cambridge, market finance was started at the height of the crash in 2009 when small companies were being abused by large corporates again in supply chains.


So it was about broadening and democratizing access to finance, particularly when large corporates were using their SME suppliers as banks, even though they can't afford it. And I think the range of problems we solve here is vast. I think if we can eliminate terrorist financing, human trafficking, money laundering, and the whole range of crimes then was that a consideration and is it a consideration for the whole team and myself?... I think absolutely. I think to be able to use technology at scale to solve really important social problems and have a really beneficial impact on the world is a great motivator.


Alberto Lidji: Yeah, that's wonderful. And the investors themselves, when you're going to market and you're seeking funding, do they ask... do they like the fact that you're not just a good business model, but actually that you are solving these social problems or you're tackling these social problems?... Do you notice that that's a consideration for them and do you notice that that's an increasingly important consideration for them?


Charles Delingpole: Yeah, I think particularly with institutions that we have on board... Balderton, Index and OTPP, I think, they're kind of very prestigious institutions and very well known. And I think it's very important for them not to be associated with companies that are engaged in transgressions or anything bad. I think there are many examples of venture capital companies who now their policy is instantly to disassociate themselves from anyone, any company, despite the profits, because I think having high ethical standards is... particularly given, generational shifts in attitudes is critical because the best companies want ethical leadership. And therefore I think... if a venture capital firm is simply often a brand and a team of people with no physical assets. Then reputation could take a lifetime to build up, but moment to lose.


And therefore they have to protect that vigorously and therefore understanding your counter parties, both in terms of supply chains, but also in terms of who you choose to partner with is becoming more and more critical for investors every day.


Alberto Lidji: And give us an overview on the scalability of things and the scale that you've built up so far and where you're looking to take things. Give us a little bit of insight on your growth trajectory and the challenges that you've faced there and where you want to take things in terms of the scale.


And presumably the more scale you have, the more robustly you might be able to tackle some of these social problems that we're facing.


Charles Delingpole: So in terms of scale, I think there are different dimensions. I think the key thing for us is the architecture of the data. And if you want to track all 7 billion people and X billion companies in the world, then you can't do that manually.


And therefore everything has been about the scalability of the underlying dataset as in the number of facts, the number of connections the number of categories, the number of, linkages, but also the speed and objectivity over data. As in, um, you want to be able to have a live feed of constantly improving, constantly evolving datasets.


So once a fact becomes apparent, you can use things like the semantic reasoning on the knowledge graph to be able to change. Because I think particularly in data a single fact can radically alter your understanding of the risks involved. As in, if you know, this person was born on a certain date or was friends or a known associate of an entity, then that means that they could be high risk for you. So, um, everything has evolved around using increasing competitional power to make the perfect data set, which can help companies and people manage risk. In terms of the company, I think what we've done recently is invest in perhaps fewer, much more expensive engineers and leaders across the board.


I think really the data set that we have will kind of carry on evolving in sophistication. And there's a virtuous cycle between adding more people, investing further and... we saw things that were tiny compared to the trajectory that we're on.


Alberto Lidji: It seems like everything you do, all the ventures that you've done, somehow they all scale. Is that always a factor for you when you're thinking about what's going to be next?... do I want something that's going to be scalable?


Charles Delingpole: Yeah I think you know, you live once, right? So, you know, dream no small dream because no small dream can move the hearts of men.. Right. If you aim for the stars at least you hit the moon, right? So I think people who have unambitious goals will always tend to underperform them, right? So I think this was deliberately impossible to replicate as well, as in, at least once it's built there'd be no point trying to replicate it. So we have hundreds of the partners who license our data and resell these because it makes sense for it to be a single company rather than multiple companies... it's a single dynamic graph of data that's constantly evolving. But also it can be used to solve countless social, environmental, governance problems. I think deliberately it was a massive scale and therefore motivational to people, both in terms of solving ethical problems, but also in terms of the potential application of algorithms and machine learning on a new size.


Alberto Lidji: Just to give us an idea of size of the company's size. What are your revenues and financials? What does that look like?


Charles Delingpole: It terms of, does it head count now we're kind of 300 people or so and the average salaries, it's quite high and we're not burning too much cash although raised a hundred million.


So yeah, I think we're still growing quickly. We had a record quarter this quarter, a record quarter the last quarter... I think there is overall growth, but we haven't disclosed revenue recently.


Alberto Lidji: Fair enough. And your clients, what sort of entities come on board and say, yep this looks interesting, we'd like to use your services. And on top of that, do you have some NGOs perhaps, or organizations that are focusing on... not necessarily for-profit entities, but who are focusing on tackling some of these social ailments?


Charles Delingpole: So, yeah, exactly. So it's a huge range of clients from banks, payment companies, insurance companies, high-risk corporates one of the first companies we had was a remittances company sending money to Somalia Afghanistan, Pakistan. So we worked with the Department for International Development [DFID] to help broaden access to that data because the money was going back in remittances, the people's families. And if it wasn't going to their families, it was going to the Taliban, Al Shabaab, ISIS. So, that was a really important dimension. So yeah, we have charities, high-risk corporates who look at their supply chains...


Alberto Lidji: And this is not just a tool that's useful for large entities, if you have a small entrepreneurial venture and they're again, also trying to be public about their supply chain and say, look, these are the steps we're taking to mitigate the risk on ESG in our supply chain... as small entity can use your services, just like a large one.


Charles Delingpole: Exactly. And I think part of it is to build at once to everyone such that it isn't just the big companies who can have access to this kind of technology and data. And we also have a program called ComplyLaunch for startups who wouldn't be able to afford that level of scrutiny on their underlying risks. So, yeah, really we want the whole world to use what we have.


Alberto Lidji: How long did it take to put together this particular product that you're launching now? How long did it take from concept or idea to, you know, waving the flag and saying, look, here we are, we're launching?


Charles Delingpole: I love to say that we built it really quickly. But unfortunately it has taken us best part... and we started working on this in 2019, so kind of over two years ago. It's a team of 30 that have been working on this currently. I think, unfortunately it's still not finished as in this still going to be hundreds of improvements that we need to make as well. So yeah, I think building this is very expensive, but we think it will have a big impact.


Alberto Lidji: Do you find any of this stressful at all? So, and I always joke with you, you know, if somebody could see you, I don't know, you always have an incredible abundance of energy. I always think like, you know, Charlie's just drank like a two liter bottle of Coke. That's how you come across to me... always energy, always very sharp, always, you know, something to say. Do you find any of this stuff stressful or you just thrive on it?


Charles Delingpole: So it's my third company. So I think getting to like the first hundred people was super easy. I've done it in the previous two companies. I think what's challenging is kind of when things aren't working out well and you don't know why or how to solve it. The hard thing I think is getting the right people on board and managing people problems as the company evolves.


I think in the past year we brought on board a whole new team in terms of CMO CTO, COO and that's had a huge, positive impact. So I think really the biggest challenges are getting the right people to be involved in a team. And I think we have that now. So right now everything's going well, it's all kind of growing... but yeah, I think there was problems are ones that he don't know how to solve and the kind of structural issues... those are stressful.


Alberto Lidji: And in terms of the road ahead. So if we're having a coffee in 10 years' time, and I'm asking you about your successful journey, what does success look like to you for the next 10 years? What is it that you're hoping this venture of yours to transform into.


Charles Delingpole: So I think if we can eliminate terrorist financing, money laundering, most ESG problems and deepen, broaden, enhance the data set. And there are tons of potential improvements we can make. But I think the key thing for me is to have an impact you need to be at real scale. So we're tiny compared to the overall market for what we're doing. I think really it's the question of scale and size and sophistication. I think we're still at the beginning of a very long journey.


Alberto Lidji: What does the market look like? You're saying you're small relative to the market, what does the market look like for this?


Charles Delingpole: I think we have a, probably like a 2% market share. So, I think still there is a huge way to go.


Alberto Lidji: So tell me, Charlie, one of the things that I've had to deal with in the past is reputational risk. And when you started looking into that, there's different databases out there, and they'll say, yeah, we can screen out against a politically exposed persons... we can screen out for various factors... and this is also of interest to those who are listening to this show who might be involved in fundraising, because you want to make sure you understand who your donors are. And so give us a little bit of a feel for that, terms of reputational risk, politically exposed persons, sanctions lists, all of that, and how did it all comes into play?


Charles Delingpole: Great question. I think another critical dimension of risk aside from ESG risk, or at least probably similar to it is... sanctions, political exposure and adverse media. So, sanctions is asset freezes, travel bans... we saw that China has sanctioned many MPs in the UK and in the EU because of the Xinjiang and Uyghurs, genocide claims. So you have to be aware if those people are your counterparties of those restrictions. If you have donors that are giving you money linked to political exposure, as in are they exfiltrating funds from their country they've stolen.


So I think you need to know... Is the money that you're taking stolen the money from foreign governments. And for that political exposure is very, very important. And then the third category is adverse media. So that's are the allegations in the press. So I think before you get involved in these situations with these people who are going to be front page news and the ostracized, then you want to check them out and you needed a data set that's live, broad, detailed, up-to-date... and that's what ComplyAdvantage offers.

Alberto Lidji: And following up from that... so at some point, though, you must need a human to look at things, right? Because I'm just thinking in some countries, when you're talking about media coverage, Uh, media isn't necessarily the free press in every country. And in some jurisdictions, it's impossible to be anyone without having, to be subjected to some sort of smear campaign here or there. It doesn't mean that that individual isn't legit... it doesn't mean that that person is doing anything wrong, but it just means that the very fact that you're active in a specific country means that you will have the press coming after you. And then presumably then your system might flag up certain things. What do you do then to make sure that we're balancing things out and that we put things into context?


Charles Delingpole: That's a great point. I think our approach of algorithmly automating the different sources and providing you with an overview of those claims means that then you can make those judgment calls.


So I think really our job is to make you aware of the accusations, show you the provenance of the claim, and then provide you with that data ready to make decisions, rather than for us to cast aspersions in our opinion. So really it's a tool for you to help avoid risk, manage risk, understand risk, rather than for us to try and keep simultaneous tabs on 7 billion people in X billion companies.


Alberto Lidji: Right. And how are you feeling about the state of affairs in general? So is the direction of travel positive in terms of tackling money laundering of identifying unsavory characters. Do you find that the technology is moving to such an extent that actors are able to identify problems earlier on and hopefully start doing something about it?


Charles Delingpole: I think progress is very slow. I think there are still X trillion dollars a year that's laundered and terrorists have no real problem in financing their activities. So I think we still have a huge amount of work to do. I think the future has to be technology and improvements in data architecture. And I think that's where we think we'll solve a lot of these key problems.


I think societies are less and less tolerant of evil occurring and it's really for institutions and companies to get a grip on their supply chains and the risks facing them. Because if they don't the general public and regulators after that will crack down instead.


Alberto Lidji: Before I forget, what's your website address?


Charles Delingpole: We're at complyadvantage.com ...on Twitter, the web and Snapchat.

Alberto Lidji: Excellent. Before we part ways for today, is there a key takeaway you'd love to share with our audience, something that if they kept nothing else in mind, but they remembered one thing from today's insight, what would that be?


Charles Delingpole: So I'd say. A lot of very influential people listen to your podcast, Alberto. And they're in key decision making roles all over the world. And I'd say they can, by deciding who they transact with and who they do business with solve many of the key problems in society. And we'd love to help them do that.


Alberto Lidji: Excellent. Charlie. It has been an absolute pleasure having you on the do one better podcast today. And it's always great chatting with you and learning from you as well. And I wish you continued success with this venture and with the launch of your product as well. So thanks for taking the time and for sharing your insight with us today.