Startup Lamassu makes bitcoin ATMs that take in cash and exchange it on the spot for bitcoin. The company was started by Zach Harvey, Josh Harvey and Matt Whitlock just one year ago and, despite regulatory challenges in various locations, more than 200 people have bought Lamassu machines and are working to put them into operation all over the world.CoinDesk sat down with Zach Harvey at CoinSummit to discuss the machine’s roll-out, Whitlock’s departure and the bitcoin ATM’s true potential. Brothers and co-founders Zach and Josh Harvey with a Lamassu ATM.CoinDesk: How many Lamassu ATMs have you sold now?Harvey: We’ve sold over 220 machines, and we’ve shipped around 80 to 100 machines. So about 120 people are still waiting for their machines. And new orders are coming in daily.Tell me a little bit about Lamassu the company. Did you co-found Lamassu with Matt Whitlock?I co-founded it with my brother, Josh Harvey, and Matt Whitlock. Matt Whitlock was with us early on, and then I guess in August, September, we parted ways.Was it an amicable parting?Yeah. He actually came with us on, not the most recent trip to our facility in Portugal, but the one before that, after we had parted ways. He still helped us out. Whenever we need help, he helps us. He’s still a good friend. We see him at every bitcoin meetup in Manchester [New Hampshire, US], which is once a week. He’s very excited about what we were doing.I guess, you know, it didn’t work out as a team, moving forward. But yes, we’re still good friends.So, now you and your brother are running the company?Yes. I’m the CEO, he’s the president. We’re just playing different roles in the company. But we’re both co-founders. We have adjacent apartments and we spend a lot of the day planning everything. He’s more on the technical side of things, he’s a computer programmer. And I’m more on the management side of things.When did you start the company?We started the company even before we were doing the bitcoin machine – about two weeks before. We were already working on the machine as a project, we just didn’t think it was going to be a business.What we wanted to do was get into bitcoin. We were thinking of doing a bitcoin consultancy, specifically with regard to bitcoin security, trying to help some of the players get into bitcoin and handle security, which is when we founded Lamassu.Then we realized it would be one of those businesses that would probably take a lot of time to develop, would not be easy. And then when the bitcoin machine really started to take off, we just used the Lamassu branding, etc.Why had you created the machine, if not as a business?It was part of the bitcoin meetups in Manchester. It started off with Matt and Josh, and I joined later. For the first time we were going to show it, I helped a bit with the last touches of the machine. The second, we started seeing it as a business opportunity, we started off as the three of us, with Josh and I doing most of the legwork, and Matt was the lead developer.Speaking of development, I have heard from the operators that they were expecting the machine to have an administrative interface, and they were getting frustrated that they don’t have it yet. Is that still in the works?It’s still in the works. We are extremely close. We have a working alpha version that we have not yet presented to our operators, but we are very close to being at the stage where we will. It was a little bit more ambitious than we had first anticipated, and I think when it comes out people will understand why it took so long. It’s a game changer in a lot of ways.The beauty of it is, it gives our machine an additional layer of flexibility. It really turns our machine into an ‘anything bitcoin’ portal. When you think of an ATM, you think of a machine that just does one thing. But this will open our machine up to have a marketplace of different services that can be used through our machine.It kind of reminds me of when you get a router and you go and configure it. There’s a URL you type in, and all the sudden you have an interface.The operators will be able to determine what fee they wish to take, what exchanges they want to use, which wallet they want to use. Maybe they want to use BitPay’s or Bitstamp’s price ticker, maybe they want to create their own. Maybe they don’t want to use any third-party services at all. Maybe they want to create backups in case the third-party services they’re using go down.One of the things it’s going to do is, they’re going to have absolute control of their system, because it will be running on their own dedicated server.They will have a server of their own that will run this machine and will be divorced from Lamassu as a central service. We will not have a central service. Which means that even if we disappear off the face of the planet, their machine will be absolutely operational.They’re going to go and start their own server, and they’re going to load up the back end onto that, pair the machine to it, and we have absolutely no control over it from that point, and no information from them.So the interface gives them the ability to connect to a server, and it’s up to them to go get the server?They can choose any server they want – Amazon has services where they can buy part of a server, there’s another one called Digital Ocean. People who are very wary of their security will want to get a super high-security one. Or maybe they want to use their own if they have some part of a server farm.There are all kinds of different levels of security they can choose for themselves. It will be up to them. The ones that don’t really have an idea of what they’re supposed to do will take servers at cloud services that are easy to set up, and people who are experts at networking and services and security will want to handle everything themselves.What other services could they provide besides exchanging dollars for bitcoins?That entire system is going to be open source. Anybody can write a module that can plug into a system. If any of the operators want to use it, they just pull in that module. So anybody that has any sort of API can write a module for our machine, and then if the operators want to use it, it could be payment based, or not, if somebody’s doing it just for the fun of it.So it will be like writing apps for an iPhone?Right.Do you have ideas about what kind of services people might write modules to provide?I would be most excited if it were things I didn’t even think of, but the basics are, for instance, more secure wallets. There are all kinds of wallets coming out right now, and some of them will have an extra layer of security. You could plug those into our system and use them as wallets instead of using the standard ones that exist today.Or maybe compliance – maybe somebody has a really simple easy-to-use compliance system. They could write a module and plug that right into our system so the operator would have a very easy time with compliance in whatever jurisdiction.Then you can start looking at our machines as a portal based on the bitcoin platform. You put cash into this machine, and then anything could be done with that once it’s bitcoin.It could be sent as a remittance to another country. It could be used to purchase anything really, directly. You don’t even have to necessarily touch bitcoin.So if I’m someone who wants to send money overseas, I could put my dollars into the machine, and not only convert them to bitcoin, but also send it overseas?Yeah. In fact it could even all be behind the scenes. You could have somebody putting $100 into the machine, and their relative in [a developing country] picks it up in their own currency. People don’t even know it’s bitcoin, it doesn’t even interest them. What they want done is to move their money and convert it at the same time.Say they want to send $2,000 to a [developing] country. They could type in an address or put in a QR code that contains all the information they need to send it to their relative, press ‘send’, and their relative goes to a machine that spits it out, or some kind of desk or service, or maybe the service would put it into their bank account.It really opens up our machine to much more flexible usability for the operator, but the most basic part of it is that it will make it really easy to manage and run the system.

In return it’ll make it a lot easier for us so we can focus on what’s really important as far as development goes. If they want to change their commission right now, we have to change that for them. That’s a temporary situation in which we don’t want to be. It’s frustrating for them, and it’s frustrating for us.

Is there a date when the admin interface is going to be released?

We stopped giving dates. We do take deadlines seriously. The second we realized we missed one, we didn’t want to just keep saying, ‘another two weeks, another two weeks.’ That’s a little reminiscent of other companies in the bitcoin world.

It just feels like, we’re doing our best to get this under control, working day and night tirelessly, and we’re very close. Giving a date won’t help anyone at this point. All we can tell them is, this is how far we’ve gotten and this is what we have left to do, and try to give our operators as much information as possible.

Basically what happened with this, is our hardware got out so quickly and on time, and our software was running a bit behind, and this was all happening so quickly – almost as to be expected in a young startup, that things don’t run exactly on time.

It’s not something we’re happy with, it’s something that we’re dealing with. We are really making a lot of great progress, finally. We had some bad experiences with software development, and now things are a bit more under control. It’s been frustrating, but a learning experience, and I will be very excited when it finally comes out.

That will be kind of the last piece of the puzzle that’s been difficult for us. Because everything else has been going amazingly: the hardware, our customer support guys are phenomenal, our designers in Berlin are great – everybody we’ve been working with. We really feel blessed that we found such great people to work with who are doing, I would say, beyond what we expected. Software has been the one point that has been frustrating for us.

Did the delay have to do with Matt Whitlock’s departure?

It started with Matt, but it’s hard to say. I’m not going to blame it on Matt, on the other hand there were certain issues with the software while Matt was there. It’s possible that if Matt had stayed with us, certain things would have been done already.

It has definitely been the difficulty of finding a lead developer [to replace Matt] that would really fit our business needs, of getting things done fast with very good communication. It’s something we’ve been looking for for a long time. Right now we’re looking for a lead developer still, and Josh is doing most of the work.

Have you had any pushback from authorities since the Lamassu machines have been deploying?

No, we haven’t had any contact whatsoever with authorities from any country. We’ve had authorities talk about our machine, but not to us directly.

In Massachusetts, the Office of Consumer Affairs put out a consumer warning about bitcoins and bitcoin ATMs, and went on to say that our machine wasn’t a bitcoin ATM, which made me happy.

In Singapore they talked about bitcoin ATMs, they talked about our machine, kind of like a general warning. In a sense, that’s exciting. But we haven’t actually been approached by any government agencies from any countries. Part of the reason for that is that we’re not an operator. We do not provide any money services at all.

What about the operators and their experiences with trying to register with various authorities – what has their experience been?

They are all facing obstacles, especially in the US. There people are very wary of what will happen if they start operating a machine. There have been guys waiting six months because they’ve been concerned about how to launch it.

Others haven’t waited that long, have engaged their regulators and gotten pretty good ideas or at least have enough communications to feel OK about it.

In general they’re all being very, very careful in the US. Abroad they’re usually quicker to launch, but some also have some concerns, depending on the jurisdiction that they’re in.

I spoke with one attorney who warned that the regulatory structure, especially in the US, is not yet adequate for people to operate these machines in a legally safe way. How do you respond to that?

‘There’s nothing to do’ isn’t really an acceptable answer. Bitcoin is not illegal in the US. So saying, ‘There’s no regulatory framework, therefore you shouldn’t do it’ … that is probably not a lawyer I would go to.

I would go to a lawyer who says, ‘There’s no regulatory framework, let’s find it’. If you go to a regulator and he says something like that, that’s definitely not acceptable. As a regulator, you can’t just say, ‘We don’t have a framework so you can’t do it.’ That’s not according to the legal system.

If they have no idea who to approach, we can recommend a lawyer to them. Some of those will give the first hour free, just to get an idea before they really get into it. So they can call them up and say, ‘Listen, I’m thinking of putting a bitcoin machine in Arkansas. Should I, should I not? What do you think it would cost me to do it?’.

That’s the kind of way we could help them. Other ideas would be our experience, knowing where it’s easier to set up. Or if we already have an operator in some state, we can say, ‘We have this operator here, I know what they did. They have special requirements’, like a special screen, or special requirements with ID verification that other states don’t have – like an extra step.

We want it to be as easy as possible for these operators. The only thing we won’t do, is get it compliant for them. We would have zero time for anything else if we started touching regulatory compliance, and we would probably run out of money in a few days.

Will the administrative interface do the ID checking that regulators require?

It’s part of it. It’s something we’ve been working on and are making progress with as well. For us, it’s just a matter of resources and priorities, and for a while we were selling more machines to countries that didn’t really need it as much. Now that we are selling more machines to areas that do need ID verification, we’ve been putting more resources into it.

The basic way it’s going to work is pretty simple: before you scan your QR code, you scan the back of your driver’s license. It reads the bar code on the driver’s license, snatches all the information, sends it to a third party that specializes in ID verification. It asks verifying questions, in the same way that you have to answer questions through some kind of fraud alert on your credit card to make sure there’s no identity theft.

We verify and validate their identity, and then they can go and buy. It’s a process that should probably only double the amount of time it takes to use the machine. Instead of 15 seconds, maybe 30 seconds.

That’s going to be part of the interface as soon as it comes out?

Yeah. There’s going to be a compliance tab.

Do people have to pay for your machines when they order?

Yes. They pay when they order, the full amount. People started feeling a lot better about that when they saw that we are actually shipping these machines.

A lot of respect to the first 15 that ordered on blind faith. I’m happy that we delivered on time for them. We have not been getting any complaints about people getting their machines late. People have been generally very happy about that.

What percentage of customers paid in bitcoins?

Definitely most. I would say probably about 70%.

How has the fluctuating exchange rate affected your company? Has that been, in general, beneficial to you?

In general it’s been beneficial. We have to be responsible enough to make sure that, with the money we receive, we can pay our suppliers in order to build the machine.

We don’t keep everything in bitcoin. We have to make sure we have a large enough amount of fiat currency so that we are able to pay people in fiat currency.

If we’d received orders from machines when they were $1,200 and left them there, it would be a lot harder to pay for them now that bitcoin is worth half of that. We’re careful to keep all the money we use for operational uses in fiat right now. A large percentage of our profit is in bitcoin.

Are you profitable already?

Yes, we are. It’s part of what has allowed us to ramp up production, giving more money up front to our manufacturers, and making sure we have enough stock of everything and keeping enough padding so we don’t have to worry about what will happen next week.

If the price of bitcoin had remained static, would you still be profitable? Or is your profit mostly due to the rise in the bitcoin price?

Yes. Our profit is from selling the machines mostly. Whatever part of it we kept in bitcoin obviously would be more profitable. At first we kept a smaller percentage in bitcoin and, as we sold more machines, we started taking a little more profit in bitcoin. We expect the price of bitcoin to go right back up.

What is the case for having a physical ATM when anyone could simply buy bitcoins online?

There are a bunch of different ways of looking at it. One is if you pick a random person off the street, what would be easier to explain to them to do, go to a machine, put in a dollar and get bitcoin? Or explain to them how they have to sign up for this exchange – that’s not really a bank, but has to talk to their bank – buy a minimum amount, and wait a few days to actually get them? It’s a more arduous process than our machine.

If you take somebody off the street, it would be a lot more intuitive to just press a few buttons and get it. And for people who are already familiar with bitcoin, it’s an easy process.

It’s really a matter of convenience. There are use cases for exchanges and doing it online, and there are use cases for having a physical machine. Obviously, if you’re talking about the unbanked, they can’t use an online exchange. It cuts the banks out of the picture. Cash is the only way to use fiat without middlemen.

ATMsLamassuZach Harvey