Episode 127 of the Public Key podcast is here! This week we have the rare opportunity to talk to one of the visionaries who started it all and built Chainalysis to what it is today. Michael Gronager (Co-Founder & CEO, Chainalysis) shares fascinating stories about the early days of crypto exchanges and why building a blockchain analytics tool was the cornerstone in getting banks, regulators and law enforcement onboarded into the future of cryptocurrency
You can listen or subscribe now on Spotify, Apple, or Audible. Keep reading for a full preview of episode 127.
Public Key Episode 127: Cracking Crypto Crime: How Chainalysis Became Law Enforcement’s Go-To Tool
Very rarely do you get to listen to top visionaries like Elon Musk, Jeff Bezos and Sam Altman talk about technological shifts that changed the world, but in today’s episode, Ian Andrews (CMO, Chainalysis) speaks with Michael Gronager (Co-Founder & CEO, Chainalysis) about building the the first blockchain analytics company that forever will change cryptocurrency industry.
Michael takes us on his journey from chemistry and computational studies to virtual reality and launching one of the first crypto exchanges.
He shares his innovative approach to cryptography and distributed systems and how this benefited him with his time at building up Kraken and recounts the early days of cryptocurrency before regulations and bank acceptance on his path to creating Chainalysis.
The duo discuss the evolution of crypto compliance, the creation of Chainalysis’ flagship solution, Reactor, on a shoestring budget and the challenges faced in bringing transparency to the crypto world.
Quote of the episode
“The idea was then basically to say, on the blockchain, we should be able to build something even better than the personas that you build from cookies on the internet on Web2. You should be able to build personas around every wallet and understand where did it get its funds from? What was happening?” – Michael Gronager (Co-Founder & CEO, Chainalysis)
Minute-by-minute episode breakdown
2 | Michael’s journey from chemistry to computational studies to virtual reality
8 | What happened when Michael finally hears about Bitcoin in 2010
11 | Challenges and innovation in Bitcoin micro payments
17 | Launching Kraken: Overcoming challenges in early crypto exchange era
24 | The industry tipping point that started the creation of Chainalysis
32 | The serendipitous journey of the founding team of Chainalysis and first investor / customer convos
39 | Crypto is starting to look more like finance rather than finance looking more like crypto
42 | The evolution of trust and stablecoins in the crypto ecosystem
Related resources
Check out more resources provided by Chainalysis that perfectly complement this episode of the Public Key.
Speakers on today’s episode
Mentioned Episodes:
Ep. 11: Chainalysis: The Past, The Present, The Future
In this episode, we talk about going from Bitcoin skeptic to believer, with Chainalysis co-founders Jonathan Levin and Michael Gronager.
This website may contain links to third-party sites that are not under the control of Chainalysis, Inc. or its affiliates (collectively “Chainalysis”). Access to such information does not imply association with, endorsement of, approval of, or recommendation by Chainalysis of the site or its operators, and Chainalysis is not responsible for the products, services, or other content hosted therein.
Our podcasts are for informational purposes only, and are not intended to provide legal, tax, financial, or investment advice. Listeners should consult their own advisors before making these types of decisions. Chainalysis has no responsibility or liability for any decision made or any other acts or omissions in connection with your use of this material.
Chainalysis does not guarantee or warrant the accuracy, completeness, timeliness, suitability or validity of the information in any particular podcast and will not be responsible for any claim attributable to errors, omissions, or other inaccuracies of any part of such material.
Unless stated otherwise, reference to any specific product or entity does not constitute an endorsement or recommendation by Chainalysis. The views expressed by guests are their own and their appearance on the program does not imply an endorsement of them or any entity they represent. Views and opinions expressed by Chainalysis employees are those of the employees and do not necessarily reflect the views of the company.
Transcript
Ian:
Hey, everyone. Welcome to another episode of Public Key. This is your host, Ian Andrews. Today I’m joined by my friend, colleague, my boss, Michael Gronager, CEO of Chainalysis. Michael, welcome to the show.
Michael:
Thanks, Ian. This is fun. I look forward to talk to you today.
Ian:
We did one of these episodes way back in the beginning of Public Key. I think we were actually at one of the LINKs Conferences and we grabbed you and Jonathan, pulled you in, did kind of an ad hoc recording, but I feel like we never got the chance to really go deep into your personal story, some of the early years of Chainalysis. So I’ve been dying to do this episode. Super excited to have you here today.
Maybe you can take us way back in time. Early in your career you studied at university, definitely not blockchain, not crypto. Take us back to that point in time and how you found yourself in graduate advanced science studies.
Michael:
Yeah, so I think I’m always been curious to understand how things work to put it that way, like what are the rules of the world. And I think that took me first to chemistry. I actually think that to be completely honest, I found physics more interesting, but the people who had chosen physics were less interesting than the people who chose chemistry. It’s completely honest. And then I basically diverted towards chemistry.
But when I then started my studies, I kind of turned it a lot towards theoretical physics, which were like quantum mechanics applied to molecules with a laser playing around with it as well, and then collaborated with a group at another university around experiments and so on. And it became why I did the theoretical part. I did the computational studies as well. Computation was a challenge back then because to get to the scale and size of computers were not that easy. So distributed systems became an important part of solving some of those problems, and I felt that was kind of the theme going forward from there, that the distributed systems became a big part of my career.
Ian:
Well, and you were doing your studies right around the time that the internet was becoming something known outside of the university framework, right? This was sort of the early to mid-90s as companies like Netscape were bringing the web to the world. What did distributed systems look like as you were completing your PhD?
Michael:
Yeah, so I remember, I think I was probably still bachelor’s student at that time where we had this Polish postdoc that showed me first Linux and then later Mosaic. Mosaic was the first browser and we were super excited about it. Then me and one of the other, I think, grad students who were at that time, we started to set up a webpage for the institute and was presenting what we were doing, doing small movies. Everything became like, how do you present something online and how do you first love the marquee letters blinking in funny colors and then later become super embarrassed about them. There’s a lot of parallels to the NFTs today, I think. But definitely seeing some of that in the early days.
Distributed systems back then, well, the internet was part of that, right? Because it was pretty hard to build without a network, otherwise there’s not really any distribution. But one of the things that I learned pretty quickly was that you could connect all the, I would say, workstations into a little supercomputer. And for the right problems, you would basically be able to solve it that way. So you could use your university account and install various distributed systems on them and then make them work together. And that was super exciting. So that was one of the first things I did there.
Then you started to use more like purpose-built supercomputers where you have better internet connections between them. Interconnect becomes really important for most problems unless they’re trivially scalable. And then later on when I left my PhD, I got involved in virtual reality. Virtual reality was also like, if you want to run multiple displays and multiple systems, you can buy a very big supercomputer from Silicon Graphics back then that would cost you millions of dollars or you could use NVIDIA graphic cards and buy a systems together. And we then built some of the early systems for that and actually ended up building a little company around that, selling these solutions to national broadcasts to other places where we could show 3D models on multiple screen display in a distributed system all synced up with home written software. So that was the earliest version of, I would say, distributed systems.
Ian:
The Metaverse 0.1, maybe.
Michael:
It was the Metaverse 0.1. And trust me, it was fun to see a virtual reality conference in 1999 where you would have people who were building things that we are not even there yet today. So shared experiences in the virtual world, but also I have to say back then, I was a skeptic in virtual reality even being super deep in it, mainly because of headache, I would say, because it’s just… I know you also have the Apple Vision Pro. I have one as well. I used it for a week, and that was basically it because it simply gives me headache. It’s just how it is. I feel that I might not prepare percentative, but a good chunk of people probably get headaches from those as well. We don’t get headaches from phones, that’s pretty obvious, because people tend to use them all the time. So I think that there’s some challenges in the interface that haven’t been solved yet.
Ian:
Yeah, absolutely. Now, many people that take the educational track you took through a PhD program end up lecturing in university, but you went pretty quickly out into the real world even though you were still in research, but you were actually not just teaching in a university doing research. Talk a little bit about that period of time through the 2000s.
Michael:
Yes, it was kind of interesting because what happened after the virtual reality, actually 9/11 was a core part of all of the things that happened there because before 9/11 there was the dotcom was going crazy and there was a lot of optimism about computers and how they could be used. Virtual reality was part of that. But then 9/11 happened. And I was actually in Manchester at a virtual reality conference. We got stuck there for a little while because planes were down. And when we started flying, there was this most silent plane I’ve ever been at and we just watched sitting live in the airport and see the plane crash into the buildings in New York, which was absolutely awful. So that was one of the worst experiences.
But it also meant that the world changed overnight, and it meant that a lot of the activities got diverted into international security. And it also meant that I actually started after that at the [inaudible 00:07:25] Institute in Denmark and ran their research programs, mainly big EU projects about how do you build big computer centers in a distributed way. And that became, again, it was grid computing as it was called, which was the predecessor of a cloud, and it was really about how do you make a big computer out of a lot of smaller ones and able to build an economy around it.
So as it sounds, a little bit like crypto in many ways because you’re trying to schedule jobs on a big computer that’s not really up all the time and you try to make an interface out of that. And we spent a lot of time on that. We even discussed how to build money systems on it, but never figured out how to do it. So it was just to put it, that was some of the early things.
But one of the things that was important there was cryptography. So I got deep into the cryptography route, how do we build it, what are we doing there, and I ended up lecturing in that at university in how do you cryptography, how does it work? And so on. So I had kind of started to get some of the ingredients, not knowingly, to make the innovation of crypto really relevant for me. So at least I was super excited for that because I had cryptography, I had distributed systems as well. I didn’t know anything about money, but that was another side.
Ian:
Well, when you finally heard about Bitcoin, having the underlying foundation of cryptography distributed systems must have been a huge advantage. I mean, I think that’s more than most people that have heard of Bitcoin to be able to understand those two key building blocks.
Michael:
Yeah, but it was because it was super exciting. I remember I had heard about it on Slashdot in 2010 and was just yet another project, so I hadn’t really paid attention. And then I read an article about Mt. Gox being hacked at the Bitcoin price tanking and going down to $2 or $3 in the summer of 2011. And then I was like, “Oh, something came out of it. This is super exciting,” and it’s a project. And just the fact that there was enough activity around it so people bothered writing and there were some people who had lost money, that was kind of great as a symptom of like, that there was interest and something was going on. So then I downloaded the client. I remember the first thing I did on my laptop was press mine, and then I was waiting a few minutes and realizing that didn’t really work because it was too late for doing that on a laptop. So no, Bitcoin magically appeared there.
But then I started to go deep in the source code and I basically read through every single line and trying to understand every single detail of it to figure out whether it actually worked or not and what was actually the core assumptions here. You go deep into that in all corners from like, “Does hashing the algorithm, is that actually as foolproof as we think it is? Is there a quantum problem in quantum computing if we have that suddenly and all of these things?” and realized that it was actually a pretty healthy design, all in all. The main risk is basically that miners suddenly are disincentivized from mining. And we’ve seen that in various, we call them [inaudible 00:10:32] T-coins. In back then, that was kind of the word for them. And it’s like one of these things where suddenly there’s no miners bothering to mine a coin and then the rate just drops, so you have a transaction every two days. And it doesn’t really work because that’s the block time.
That’s a challenge and it will adjust itself, but that will take years and that incentivizes further. So a lot of the blockchains ended up picking back another blockchains. So they had like, you could mine a Bitcoin block and still get a… What was that? In the bigger early days, both Litecoin and Namecoin and a few others who basically could inject something in a transaction and then you could mine for many chains at the same time. So these things were solved, but you look deep into that, but that was the main problem. That was really like if the interest fell over a certain point, then it would kind of die in an oddly way and people would need to have some intervention where they would fix that.
Ian:
So from that exploration where you’re sifting through the source code and you get to this point of, “Wow, this actually works. There’s no obvious broken pieces.” As long as there’s enough people interested in it, it’s likely to continue,” what did that mean for you from a career perspective? Was it immediately like, “Oh, I want to be working in and around this technology”? Or was it still more like a hobby? “This is interesting technology that I want to keep playing with, but I’ve also got my day job.”
Michael:
Yeah. So I kind of didn’t have a day job at that time because I was like, at the same time I had left my career at [inaudible 00:12:08] Institute at what was related to that. And then I had started to code iPhone apps a year earlier, and it was more like to try something completely different. If you can call it the midlife crisis, whatever, but it was definitely saying I wanted to try something else out. And at that time, there was no one who code apps. I couldn’t of course either, and I hadn’t been coding for 10 years, but I got back into it and got a gig with a company importing their app from Windows CE to the app store and to Apple, and it was kind of fun.
Suddenly, I had a career where I earned pretty well on doing so compared to what you get paid in the public sector. So I had a lot of time on my hand to suddenly look into things, and I decided to go deep into the route of Bitcoin. I basically decided that probably the summer of 2011, and that was just where I would say for all practical purposes, I spent full time on that since then.
Ian:
Wow. What did that actually entail? So summer of 2011, you’re all in on Bitcoin, what were you actually doing?
Michael:
So what were I doing? First, I was trying to figure out how do you actually get Bitcoin because mining them was not really a good idea. So the first thing I did was buying a few bitcoin on Mt. Gox. Then later I went to a conference in New York, a conference. We were like 40 people in a hotel room basically discussing what it was, going to cafes that promised to sell drinks for Bitcoin or coffee for Bitcoin, and we tried to system out. So that was kind of geeking around about that.
And I think that my idea was basically the world is going mobile, so this needs to work in a mobile setting. And yes, I could compile Bitcoin and run it on an iPhone, but basically it was using all the system power and it was not really… You would need a lot of all the transactions being sent back and forth. So the first thing was basically, how do you build a server client infrastructure out of this? And then the server client infrastructure became a first idea into micropayments as well, because how do you build the micropayment infrastructure? And back then, the iframe concept in web was pretty dominant. You would have iframes everywhere and you could do these things.
So I built an iframe in JavaScript that put a CD-R wallet that was living on your device. And if you didn’t have any money on it, then you were sent to a credit card page where you could buy Bitcoin and you could structure it in such a way because there was 3D secure that you could get a cryptographic proof that you actually had been sent the money, and you could do that all the way. I discussed it with the payment providers and they just did not know what I was rambling about because they’re like, “What are you even saying?” And I was basically trying to convince them to support this and not to shut it down if I were to start it. And I couldn’t get any firm connection with anyone that wanted to have that conversation, especially not in Europe at that time because that was just not a thing.
But it was actually cryptographically proven that the transaction had taken place, you had the knowledge, all of that. So there would be no doubt in the transaction even though it was a debit card transaction. But anyway, that was the way to fill up your wallet, and just like a hundred dollars or so, and then you would buy articles and other items. And the idea was basically to build an app store that was federated all over the internet. So the App Store had just emerged and been a huge success for micropayments to buying apps, to buy songs and other things. But what if that was something that were everywhere? So you didn’t have locked into one single system, but you certainly had a payment system where everyone could build things, from video clips, from everything else. So like a social media federated system where everyone had the ability to do payments and receive them.
But I also realized that if that was successful, then having around 1,000 to 2,000 transactions every 10 minutes would definitely not be enough on a worldwide scale. So I started to talk more to the code developers that I knew for a while at that time around the payment protocol and how to speed up the block size. And then the block wall was kind of slowly started, right? Because I had made a very strong proof that basically the system was designed to be self-regulatory, so regulating. So you shouldn’t put a block limit on a block because you basically compete against how fast you can transmit it, and that will basically decide the system in itself.
There was just not enough mathematical understanding from a lot of the developers. They were really good at… They were not trained, they were just really good coders, many of them. So it was really hard to sell that idea and do other things and I just felt that I was bouncing against the wall with ideas, but no one would actually be able to assist them with a proper argument.
I did a lot of built things also to scale the protocol. I built up the first Merkle 3 that was cryptographically proven and other things. I later learned that there was a price out for that, but I never claimed it. So there was different things. So there was a lot of things going on there early on, which was really fun. But tit was then from actually the first conference in New York I had met this other guy, Jesse Powell, at a breakfast cafe, and we had just stayed in contact, like, “What are you doing? What am I doing? What are your ideas?” And he had reached out to me and be like, “I’m trying to build this exchange” and whether I wanted to join. So that became the first endeavor into crypto to be part of that team.
Ian:
It’s kind of amazing the payment story, because it feels like we’re still trying to solve that today. And I’ve heard interviews with Mark Andreessen where he’s kind of lamented that not including payments into some of the early web specifications or figuring out how to make it a native part of the browser is the one thing that he would really go back and change if he could rewind time back to that mid-90s period. So we’ve got at least two and a half decades, now maybe three decades of people who see that as an obvious problem, but yet we’re still lacking a solution on the payment side.
Michael:
We do. And I would say it’s not been bad, but it’s just like the experience is, below respect, a bit clumsy, and if your browser system dies and so on, it’s okay that you have this living on your whatever profile, so yeah.
Ian:
Yeah. I think the closest experience I’ve gotten to it is maybe what Shopify’s done, where they’ve integrated the common e-commerce payments, API, that knows who I am and it aggregates up all the shipping details and all the other information into a single place. They’ve done a very clean integration.
Michael:
They do. Exactly.
Ian:
But even still, I wouldn’t use Shop Pay to buy a single song or an article that someone self-published on their website. So we still haven’t gotten down to that microscale in terms of payment volume. So somebody out there needs to go and solve this problem is basically-
Michael:
I totally agree.
Ian:
It may not be us, but somebody.
Michael:
Somebody should, yeah.
Ian:
So you and Jesse Powell get together, and Jesse propositions you and says, “Hey Michael, I’m going to launch this crypto exchange.” This is 2014-ish timeframe?
Michael:
Yeah, let’s see. He reaches out in 2012 or ’13, I can’t remember exactly when, but late ’12 or early ’13. That’s where we start the conversations about it.
Ian:
What’s the landscape of crypto at that point? Other exchanges, do they exist? Obviously Mt. Gox was around. Coinbase, I think, was something at that point maybe, but…
Michael:
I think Coinbase started, but Coinbase was not an exchange. They were broker.
Ian:
Yeah.
Michael:
They were born as a broker and they were using exchanges like Bitstamp and others to basically get liquidity from. We launched first as Kraken as an exchange. I think that the landscape was… And honestly, one of the main rationales to launch Kraken was really the concern about the dependence on Mt. Gox and the concern that it was built in a way where we felt that it was simply not stable enough and simply not good enough and at a big risk. It turned out to be true, but not in the way that we imagined. But I would say that it was definitely a concern for that and saying, “We need to build something that’s way more solid.” And then at the same time, solve in a better way the banking relationships, some of the other myths that has been going on there. So that was the main drivers for that. And then thinking that long term, there’ll be a lot of things to be traded on an exchange that would make it useful to have.
Ian:
Yeah, and what was the experience launching a new exchange in crypto? Did people immediately show up and go like, “Oh, I’ve been waiting for this to exist and I am so glad you’re here”?
Michael:
We kind of had a joke back then that there was 100,000 crypto users in the world and they all had an account on all the changes. So everyone was excited. So everyone would be like, sign up, do this and that. And I think that’s a little bit retrospectively that I think that roughly happened, that literally everyone in crypto had an account everywhere because you would do that, you were just excited about it. I think it’s the same as a lot of us geeks from back in the ’90s. We would have an account on every social media platform on everything, right?
Ian:
Sure. Yeah.
Michael:
So it’s just there. I might not have it today and bothered today to do it, but back in the early days, I would have my name on every single platform, right? That’s just how it is. So I think you did the same in crypto. You signed up everywhere. The process was relatively smooth before, all of the KYC stuff really kicked in to put it that way. So it was a little bit easier.
Mt. Gox, I should say, actually had, to my knowledge, one of the best systems because if you trade it for more than a thousand dollars, they would send you a YubiKey, which means that if they send you a YubiKey, they know where you live and they have actually a proof now that you got it. So this is the best proof of address you can possibly dream up, and then you at the same time create a strong connection to the customer because they now have a YubiKey. So I think that was awesome. Of course it doesn’t scale for lower amounts of money, but again, I don’t think it scales that well to onboard customers who will not spend a thousand dollars on your platform, especially if the rationale is that you need to know what a market order is. So I’m like, it’s just where you have that limit as well.
But I think the challenge in launching the exchange, really customers were excited about it. They loved the UI, it looked beautiful, all of that, and it did. And then we just needed a bank account. And honestly, that was one of the forgotten pieces where it was way harder to get a bank account than any of us expected. It became a little bit of, I would say hard summer, hard summer of 2013 to actually build our bank relationships. You don’t earn any money on having an exchange in better. So it was like people could play, they traded, but they couldn’t trade. And there was not enough other coins to make it a pure crypto exchange. It wouldn’t make sense alone. So really getting our first banking relationship with a bank in Europe setup was amazing, and that enabled us to now go all in and became pretty fast, the biggest exchange between Bitcoin and Euro worldwide.
Ian:
And the reason for-
Michael:
It might still be.
Ian:
The reason for having the bank account was so customers could transfer dollars or euros in and then be able to exchange that on your platform for Bitcoin or some other asset?
Michael:
Exactly. Exactly.
Ian:
What was the conversation like when you phoned up the bank and said, “Hey, we want to be able to accept wire transfers from customers all over the place and we’re going to allow them to buy and sell cryptocurrency with that”? Did people even know what you were talking about?
Michael:
Well, so it was kind of the same story everywhere. So first it’s like, “Oh, that’s interesting.” And honestly, it’s also like, if you’re in Europe and you come from a San Francisco-based company, everyone is opening the door. “This is cool.” They get royal visit in many ways, it feels like. So you get a really good treatment and they’re like, “This is super cool. It sounds like there’s opportunity here.” So everything feels right in the first and maybe second conversation. And then you get the third meeting and then they’re like, “We talk to compliance and they don’t like it. You don’t know the written of funds, your onboarding setup. It’s not like what we need and you can’t really do it.” So basically stranded there, both in the US and Europe for that matter.
Many times we were done an integration with a banking provider, like technical integration, and then they pulled the plug well after that because they just hadn’t really thought it through what was actually the proposition around it and they got concerned. And they maybe had talked to the regulator and other things, which have been like, “Yeah, it’s kind of on you.” Because in Europe it was interesting. Crypto was not regulated at all. So there were literally no rules, meaning that you also could onboard anyone and you could do anything you wanted. You were not subject to the anti-money laundering rules in any shape or form because you were just trading glass pearls. And it was not even an example of the list. It was not fine art, it was not anything else. So it made banks the de facto regulator.
And further on, the other problem were, if you then as an exchange saw suspicious activity, you had nowhere to report it because you were not regulated. So you could of course basically go to law enforcement with your customer list and the list of customers, but that’s not what you do in finance. And you just have a suspicion, you don’t have proof, right? So it doesn’t work. And that made Europe a pretty complicated place to operate. But if you had a bank account, you are basically safe because that was the best things you could have, so yeah.
Ian:
So in order to get that first bank account, did you have to do a huge amount of work on the compliance side to fill some of the gaps they were concerned about in terms of KYC and onboarding?
Michael:
We did some of that, but we kind of felt we had that covered. When I look back at things today, I would say there’s many things that were maybe not as well understood. And the main thing that became a problem in other bank accounts were really transaction monitoring and understanding the original funds. Because what’s usually say, what I realized in the early days of being part of the industry, and as I said, I do a lot of our tech and cryptography and distributed data, but not a whole lot of amount of money, I think from coming from tech and physics for that matter, you always have a slight arrogance that everything else is just another version of tech. So is there really another discipline that can’t just be techi-fied and you realize that finance act is.
So as soon as you do certain things, for example, there’s a reason why on your Amex you can’t just use it for personal payments into Venmo only up to a certain limit, because as soon as they do that, they need to change their compliance program and need to do a lot more betting. And Bitcoin is essentially that, because suddenly you buy things that way. So you need another level of understanding of your customer, and a passport is not good enough. You actually need to be able to understand your customers real time. And I think that was, as I usually say, it’s like we tried to onboard billions to transact millions in minutes. And in finance, that’s not a good idea because you simply don’t have the tools available back then to actually bid out the transactions in such a way that we do this in a way where you didn’t became a risk for the financial system.
I think today you probably can, because the realtime oversight we have of the blockchain actually enabled that over time to make that dream come true. But I think it took a lot of time and it created building another company for that.
Ian:
Well, and that’s a perfect setup for my next question, which is, it feels like you were perfectly positioned at Kraken to found Chainalysis in a way that maybe no one else was, both your background prior to Kraken, but then the experience you had there trying to open an account and dealing with this knowledge of what your customers were doing. Take us through that inspiration and the moment where you realized there’s a problem that must be solved here and what led you to actually deciding to create Chainalysis.
Michael:
Yeah. So I think I started with civil, like reading some of the, how should I put it, the questionnaire or the examples of how to build an email policy. I read that in Japanese. I can’t speak Japanese, but we got a machine translated. I had a good colleague and we read through that one. All of them were the same, whether it was Europe bill, it was Japan, whether it was in the US.
I remember one of the ideas that was mentioned there, how do you do monitor for suspicious transactions if you are a teller in a bank? Well, one example could be that some person, two people outside the door exchange a paperback with the other one, and then they run into the bank with the paperback of cash and want to deposit it. That’s suspicious activity. And then I was like, “Well, that’s hard to do online, but actually we can see around corners online so they don’t even need to exchange the paperback in front of the door because we can now go back in time.”
And the idea was then basically to say, on the blockchain, we should be able to build something even better than the personas that you built from cookies on the internet on web 2. You should be able to build personas around every wallet and understand where did they get this funds from? What was happening? Whether the equivalent of a brown paperback will cash be in exchange for something we don’t know what was. Or we might even knew that it was for something bad or we might knew it was from something okay. But it just meant that you could create the equivalent of a paper trail that you would have to ask for in a bank, but create that trail through the breadcrumbs of the blockchain and create that persona. And that was kind of the idea from, I would say, reading some of these programs and then thinking through how the blockchain worked. And then I was like, “This would probably be a good thing to do because that would solve the question we constantly have gotten from banks.” So then it was just a matter of doing it.
Ian:
It’s an amazing insight that it feels like so few people were positioned to do at the moment that you came upon it. How did you go about actually starting the company? Did you ever think about building that capability inside of Kraken? Would it have just been maybe an internal system or a feature?
Michael:
Yeah, I would say I was thinking about something like that, because we also had this conversation will build banks and I tried to teach them about that it was possible in general, but they would just say, “Yeah, but it’s not done.” And honestly, when you’re running an existing business, to do something as different as that, it’s starting a new company inside the other one. You would have 10 competing things. Should you build a broker? Which instead? Should you do this or that? Will the problem just go wave by itself? What will actually happen? Will it give you a huge advantage?
And at the same time, to be honest about it, you also know that it’s a data company. So you actually want to be separate from a financial institution, and you want to be able to work with all financial institutions. So for me it was a thought that I had, but at the same time being like, “I think this is an opportunity to try something on my own.” And if we go back to the beginning, I was like, “I should not be in finance.” I was not meant for that. I should run a data company instead. So it was just like, again, I felt that I was not doing the thing that I was kind of best at in that sense. So also leaving and then doing another company felt like the right thing to do.
Ian:
Yeah. And so at some point, I imagine, did you already know your co-founder Jan at this point?
Michael:
Yeah, so I know Jan. We’ve been comparing notes at conferences and would meet at various occasions and so on and basically discussing things. What happens there is that I leave Kraken. I’m then in Copenhagen at that time. On the trip back from when I leave Kraken, it’s a long flight, it’s 11 hours in a plane. I actually code the prototype of Chainalysis, which basically just like a [inaudible 00:33:47] algorithm with a little bit of extra on top of it. Because I needed to serve a client infrastructure for the other projects, already had a crypto server that would index the blockchain. So already had that capability. So to add the clustering on top of that was literally just another table in a database or a couple of tables and that was it. So that was pretty fast. And then after half a day it had synced up and then I could start to do queries and understand, “Oh, that service might be Bitstamp. That one must be that. That must be that.” So I had basically built that.
And then I meet Jan a couple of weeks after that where I was trying. And honestly I was still torn whether I should do this or whether I should just maybe go back and do some other job or whatever I should do because it was not that easy to decide. And then I meet Jan, and Jan was like, “Oh, we have a crap job from you at Kraken.” And I was like, “I just left.”
“What are you doing then?” And he’s like, because he was about to leave the company, he was at Mycelium. And then I was like, “I’m building this project.”
He’s, “Do you need a co-founder?”
I was like, “Yeah, sure. It’d be a good idea.”
Jan had this stellar argument that I had to be the CEO because I look better in a suit than he did. I don’t know what he actually got me tricked into back then, but apparently that was the argument.
Ian:
He’s not wrong. You do look great in a suit.
Michael:
So we basically started the company and coded and do a lot of things for the first couple of months and built an infrastructure that was built at the infrastructure option analysis in a couple of months. Got a few contractors to do an interface that became reactor. The price for that, it was actually amazing what you could do when you were just working in a scrappy way. It took them a week, one and a half week to code it, to be honest, where you were clicking around and connected to our backend.
I think we paid 1,000 or $2,000 for that, maybe it was Euro, but still.
Ian:
[inaudible 00:35:50].
Michael:
That was something along the lines.
Ian:
Yeah.
Michael:
So it was super cheap, right? So that’s basically what we did. That was the early prototype. And then from there, it only of course got harder to build from there. But I would say that was the early days. And that became the company. And then you’re probably going to ask that question, so I’m just going straight into it because the funny thing there is that we now actually have a product and I can showcase it and do other things and customers are kind of excited, including some law enforcement customers that sees it. Katie Horn actually is the first law enforcement customer we see. That’s roughly the same time I start to talk to Jonathan, the third co-founder.
Ian:
And Katie Horn, for people that don’t know, was a prosecutor with the US Department of Justice involved in Silk Road and Mt. Gox, I believe. She was one of the first people to really take on crypto related crime. So when you started showing the product to people like Katie, what was the reaction? Was it immediately, “Oh, I get this. This is amazing?” Or was it like, “Whoa, whoa, whoa. What’s going on here?”
Michael:
No, they kind of got it right because the first meeting I had with Katie and [inaudible 00:37:05] was like they had basically scribbles everywhere on the whiteboards with crypto addresses and had been looking things up in block explorers and putting in spreadsheets and trying to create their own little tracing tool that way. They didn’t really have good attributions of what things were and a good map of the blockchain, and that was basically what I could offer them. And at the same time, I could offer them the ability to then have an investigation and not just have a spreadsheet that was not connected to any intelligence real time. So that was basically reactant, the ability to store your investigation and keep it and understand it and build from that. So that was kind of our first, how should I put it, document object that was really important. And they loved it. That was one of the things there. And they loved it enough to do introductions to other agencies, which was awesome.
Ian:
And so even from that early prototype, you felt like you had some product market fit. There was a real buyer out there and some need for the technology.
Michael:
I felt so. I would say there was a buyer out there, but again, I would say the view from every law enforcement I spoke to was not that, “This is our first crypto case. It was our last crypto case, because this is going away tomorrow. We all know that. So we just need to solve this case here, and is it even worth it?” So it also meant that the willingness to invest on other things were not really there in early probably 2015. So they would buy it for shorter while, for a few months to solve one case, and then assuming that after that, that was it. And that was kind of the rationale for most things. But obviously that was not how things played out.
Ian:
Yeah. No one could imagine in 2015 that crypto would become as big as it is today. It is even hard for me to believe, and I’m in it every day all day long for the last almost four years.
So you started to mention Jonathan Levin, obviously your other co-founder. Talk about how Jonathan got involved and what that was that story in the early days.
Michael:
Yeah, so there’s actually two stories, and the first one is funny because the day after I met Jan for beer in Copenhagen, Jan talks to Thomas Bloomer, who is an early Bitcoiner as well, and Thomas talks to Jonathan, and Jonathan hears about this project we are doing. He’s already been looking at doing compliance in crypto, but in a small startup and it didn’t work out, and then he was looking for the next thing to do. And then he heard about someone doing something similar and wanted to reach out. So he sent some mail to Jan. Jan asked me, like, “Someone is reaching out” and I say like, “Yeah, I don’t know. It’s too early. We don’t really know what we are doing yet, so it doesn’t really make sense to bring other people in.” And it was kind of what it was, right? So we were just ignore… Well, we replied, but that became it.
And then a few months later, I think we in February, I guess, something around that time, maybe Christmas, maybe some other time, but at least I have a common friend in Brian Fabian from Epicenter Bitcoin. Jonathan knew him, I knew him. And I’m talking to him at a conference in Berlin and he’s saying like, “You should talk to this guy, Jonathan, because he was saying the same things and you should definitely meet.” And I have no connections between the email from back in the days, I couldn’t remember it. And then we set up a call and then I’m thinking, “Okay, you have a lot of calls and most of the times you just talk to people about what you’re doing, but rarely it becomes anything, right?” Sometimes it’s, I’ve been talking to bankers and others and they’re super excited about listening to what you’re doing, but then nothing really comes out of it and you spend a lot of time doing these calls, and I have the same expectations here.
But then it’s fun because Jonathan kind of knows what knew what we were doing in understanding that he’s seen some of the same things in his earlier company and trying to figure out like, “Okay,” and we both realized we can actually do things. And he realized we have tech, and they didn’t have tech in his company. We have built that. All of it is there. And then it became very interesting conversation. I think a week or two after we meet up all three of us in Copenhagen. And then we have this like, I think it’s very typical for the early startup phase where you have this dance. I tried to became part of Kraken, but it was a dance. It was not like it’s one off and then you start and that’s how it is. It’s more like you’re trying to figure out, “Well, that’s the right next move you should do and what’s going on.”
So we start to do more and more often calls and comparing notes and we start to discuss pricing. Jonathan had this big trip for China coming up for three weeks or so, and we kind of feels that we are like, “Yeah, he’s joining,” but we just haven’t really done the final work of he’s actually joining. And when he comes back from China, we just met with FBI and they became like they wanted to buy and now we’re like, “This is happening.” So that’s where he joins and we become the founding team together.
Basically, I remember we sitting in a little office in Copenhagen and writing all our customers on the list, and we have a lot suddenly. It’s like in a few months we built a customer list and we have to try to figure out who to reach out to. So we basically built salesforce on a whiteboard, right? Salesforce on a whiteboard, and trying to figure out what to do. None of us knew anything about salesforce or doing all of what we were doing there. And the sales motion were that we got an inbound from some law enforcement that had seen our tool from FBI from another US agency and they wanted to buy, or we reached out to a crypto business and say, “Don’t you want to do this compliance-wise? At least you know what you’re doing then.” And it works pretty well. So we built pretty fast revenue. And at the same time we are sitting there in the office, we are also realizing that we probably should move to the US.
Ian:
Oh, interesting.
Michael:
Jonathan had gotten a question from a friend of his that are in the back of Techstars program are trying to find companies for that. And they’re asking if you know any interesting crypto companies that wants to do New York. And then we look at each other and be like, “But maybe we should apply.” But the time is kind of ’01. We are both going traveling to different customer meetings and conferences and so on.
So we managed to… I think we do half of a demo. Jonathan does sitting on a floor at a conference where there’s noise in the background and I’m doing a presentation online while I have family visits for my birthday or something, and then straight into a plane and then going to the textile office and then we get accepted into the program.
So then we became start set up in New York, a part of the textile accelerator. I think that shapes us a lot because we basically then become part of another group of people who want to do the same. We can compare what are they doing, what are the revenues, what are our approach, and we just constantly talk about what is the proposition of geneticists. Why are we doing this? What is it? What is the setup? And we become sharper and sharper on what we’re doing over that summer.
Ian:
It’s such an amazing, the number of coincidences that had to work out, and that you almost didn’t end up in the Techstars program, because prior to that, you’re building in Copenhagen. It sounds like you’ve had some interactions with a number of companies all around the world in the US. Had you tried to raise capital outside of the Techstars path?
Michael:
Well, yeah, we had. First of all, I don’t think we were good at raising capital first. We didn’t have a good idea of what to do and what to promise and what to do other things. And again, I remember doing probably… Maybe it was same day, so I can’t remember whether this was the first pitch, but I actually piece Excel back in probably March 2015. They don’t really bite and they’re kind of not really… They’re just, “That was interesting,” but they didn’t really get further from that.
And I kind of get it today. If I look back at that, it’s like a previous gravity. We have some traction, what’s going on. But if you compare it to, and that’s with all due respect, if you compare to all the crypto projects out there that get capital, we were far more advanced. We actually had customers, we actually had revenue. We actually had a product solving a use case. Was the TAM small back then? Yes. But if you look at many of the use cases today, it’s like it’s all standing on the shoulders of what’s being built in crypto. So you don’t really know how big it’s going to be.
Ian:
Yeah. I mean, I would imagine that was probably the primary objection, just like you said that many of the early customers said, “Hey, this might be our last crypto investigation ever, you had investors who were going, “Hey, you guys seem really bright and you built a great product that people are using, but it’s hard for us to imagine how big can this thing get.” It obviously paid off great outcome for the early people that took a bet like Sarah Tavel at Benchmark and obviously Techstars, and Philippe [inaudible 00:46:25], who believed early, ultimately, maybe not in 2015, but he came around by a few years later.
I want to maybe take the conversation in a little bit different direction. And just kind of reflecting on where we are now in 2024, 10 years on from the early days of founding Chainalysis, have things evolved in the way that you would expect? Meaning, you actually made a statement at Links Conference that crypto is starting to look more like finance rather than finance looking more like crypto. Did you have that idea 10 years ago as you were building Chainalysis, that there would be this move towards seeking more compliance and broader acceptance into the financial system? Or is that something that’s evolved over time?
Michael:
Well, I would put it this way, that when I first sold the tech, I was a techy, right? The tech of Bitcoin, what was the concept of that? I was like, “Oh, that’s cool. You can just do finance that way.” And then you’re thinking about it and you hear people saying, “Be your own bank” and so on. You’re like, “Oh, cool.” But then you can just do it.
But then the next question, you are like, “If it’s really that simple, then it should happen with much higher speed. And if it’s really that simple, why do banks even insist?” Because you’re asking yourself the question around like, “Yeah, this is a distributed database we suddenly built, which is super cool, but the banks have databases already on tech. So what are we actually then solving that they couldn’t solve already and what actually happening?”
And then I would say then my journey starts into what finance really is about, and the layers of finance and the layers of compliance and what actually safekeeps our society by safekeeping the financial system. And then I would say I pretty early on got this skepticism towards like, well, I actually think a lot of these institutions we have are there for a reason. And the reason I didn’t exactly knew what the reason was. Well, some of things are obvious, right? But I didn’t exactly know where the limit would be. What are the parts we are cutting out and solving purely true tech? And what are the parts we are cutting out and can’t cut out and that’ll continue to be institutions?
I should add, in the grid computing world, I’d been super involved in cryptography, and one of the ways we would sign into everything was through certificates. And certificates are like cryptography, you have a private key. Getting university PhDs to keep their private keys safe was a mess.
So now we are asking the world of the internet to do the same. So I was highly skeptical that he would ever have our own wallet. Yeah, I could do it. Even me, I’ll probably throw it this way as well, but I would not be mad at anyone if I did it. I would understand that it was my own fault, so that must be fine. But the world of the internet would never have treated that way. So I just didn’t believe that we were able to do this in a way that worked because I had just seen it. So I was like, “No, we will have a wallet ourself, but that will be for hundreds of dollars and not thousands of dollars. That’s how it is.”
And yes, we will have a lot of the financial institutions rebuild because they are basically the ones that safekeep it. If I get hit by a bus, who ensures that my crypto products are passed on? It’s a mess. If they were on my own phone, maybe it’s impossible to recreate them. We haven’t lived through generations yet where that’s just become a huge problem, but it will be. There are many problems that are solved by institutions, and we need them. And that’s why I was like, “This is much more email than it’s anything else.” And in email, we all, in the ’90s, run our own email server. Any institute, any like, “I had my own email server running at home with a domain.” And you would do that, that was fine. But then at some point we stopped and we ended up using providers for that.
Today we have a handful of email providers in the world, and that’s good enough. The whole strings of that system is that it’s possible for anyone to start an email server. So it’s purely federated and it doesn’t work the second that doesn’t. It’s not an option. But crypto is the same. Everyone can boot up their own node, but the second they can’t do that, the system stops working. But it doesn’t mean that everyone will boot up their own node. It means that we will have a few worldwide. A few could be a hundred, maybe a thousand, but not billions and millions. I don’t see that.
Ian:
And tie this into stablecoins for me, right? Because it’s an interesting shift, I think, from the original idea of, “We’re going to replace the existing financial system with this new distributed ledger,” a related one is sort of, “Hey, crypto is necessary because the traditional fiat currencies are just these runaway inflationary assets that can’t be trusted. So we’re going to make our own that has a different set of rules that we’re not going to allow the governments to just print new money indefinitely.” Stablecoins seem to fly in the face of that in some ways, at least philosophically, because you’re suddenly linking crypto back to the very asset that we were seeking departure from, or the original people who conceived the system maybe were. How do you reconcile that and what are your thoughts on stablecoins?
Michael:
Yeah, so it’s actually interesting. So you could say I changed, but I actually don’t think I changed my opinion over the years. But if you find an old YouTube clip of me at a Bitcoin Conference in, I think it’s in San Jose in 2015, I’m on a panel talking about colored coins. Colored coins are basically putting other things than pure native crypto things on the blockchain. And I’m saying not as we find words as I’m using right now, but I’m basically saying that that doesn’t work unless it’s inherently digital because we don’t have institutional trust building the crypto space yet.
So I can tell you that I’m going to sell you San Pellegrino tokens and I’ll redeem them, but you don’t trust me because I don’t have any institutional value that you can trust. So we couldn’t do that in 2015 because we didn’t have any… Oh, it’s actually 2014. We couldn’t actually do that, as 2013 maybe even. So we couldn’t do that because there was no institutional trust built in the space. There was no institution that would endorse crypto and this thing to such an extent that they would be willing to say, “Yeah, there’s dollars in the bank account and you will trust that. And that took time to get us there.
Today, it’s possible because we have several institutions out there that are being trusted with hundreds of billions of dollars. They are now active institutions and pillars in crypto. And we see more and more stablecoin providers popping up with even more institutional credibility than some of the ones we’ve seen in the past. And I think that that’s strong. And it basically requires institutions the trust, institutional trust, to have a colored coin. And it was not possible to do in 2015 because that was not that institutional trust present.
You could build other kind of stablecoins like MakerDAO that was like there, like Terra-Luna, like other things that pure tech, but you don’t require that piece of trust, but you couldn’t do it from the other thing. And that’s actually good. It’s not a bad thing, it’s a good thing because it means that there’s institutions in crypto today that we trust as a world.
Ian:
Well, I think that is really what Chainalysis has built the company and our mission around, is delivering that trust, right?
Michael:
It is.
Ian:
Through transparency into what’s happening in the ecosystem. So this is maybe a great place to wrap the conversation, Michael. This has been fantastic. I’ve enjoyed it greatly, the tour through how we got to this moment in the company’s history, and certainly looking forward to the future. Thank you so much for the time.
Michael:
Thank you. This was fun.
