Matt Johnson, CEO of QCWare is interviewed by Yuval Boger. Matt and Yuval talk about his advice for those starting quantum companies, how far quantum software is from being shrink-wrapped, what he learned over the past year, what QCWare will look like in 5-10 years and much more.

Transcripts

Yuval Boger: Hello, Matt. And thank you for joining me today,

Matt Johnson: Yuval, it’s great to be here. And thanks a lot for having myself and QCWare on your program.

Yuval: It’s my pleasure. So who are you, and what do you do?

Matt: My name is Matt Johnson. I am a co-founder and CEO of QCWare. And what I do is help pull our team forward so that our business can be the first quantum computing software company that can yield practical quantum advantage. That is the objective of our company.

Yuval: I’m curious about your personal journey. You’ve been doing this for more than a year or two. How did you get into quantum, and how long ago was it?

Matt: Well, I honestly truly do not like to talk much about myself. I think our team is delivering all the value. You’ve posed this question, and I’ll answer it briefly, but in a nutshell, I was originally an officer in the military, and I was always enamored with aerospace and defense technology, and just technology generally. And after that, I had gone into finance, and specifically, I spent a decade or so in principle investing, private equity. And those two professional experiences were both fascinating, and I wanted to somehow bring them together as a next chapter. And it seemed to me that working as part of a team to develop a business around a new technology, that’s the combined theme. I thought that would be very, very intriguing. And that took me to Palo Alto about eight and a half years ago to work with my fellow co-founders to set up QCWare, and we’ve been going ever since. So that takes us to today.

Yuval: We’re seeing companies that were set up eight and a half weeks ago, or eight and a half months ago. What advice do you have as CEO for someone who’s starting a quantum computing company?

Matt: That’s a great question. For quantum computing, I think that person should do a very thorough mapping of the industry. That person should study who is doing what across the technology stack, and be very sober and realistic and honest with themselves about what business prospect and technology they have that they really think is truly differentiated. And I think that exercise is more important today than it would’ve been five years ago when quantum computing companies and startups, in general, we’re getting money thrown at them. Now, we’re in a phase where all of us must deliver and will be measured as to whether our thesis about how valuable our business is, comparing that thesis to reality and you need to have a favorable conclusion or outside stakeholders have to, in order for you to attract employees and attract funding and then to attract customers.

So I think you should have a very sanguine view on this. It’s an extremely crowded field right now, and there’s tons of uncertainty around really the core kernel of what has to happen to spark, to make this industry really take off as a commercially viable thing is having a quantum computer, having hardware that is capable of delivering quantum advantage using the extant quantum algorithms out there. So absent someone coming up with an algorithm or a use case that can leverage 20 or 30 qubits and beat any classical machine out there, assuming that doesn’t happen, the world of quantum computing or enterprise computing really does need a couple of hundred very, very high-quality qubits. Perhaps not error corrected, but high quality. So that’s the spark that has to happen. And until that is the case, all the other parts of the technology stack like us as an application software vendor, we are waiting for that to happen. And it requires a lot of resourcefulness in terms of building a business, a software business when quantum hardware’s not there.

Yuval: What do you say to customers when they say, “Well, the hardware’s not there.” And I used to work for a software company, so I’ve been faced with the same questions, but what do you say?

Matt: Well, we have a couple of customer bases. So for customers who say the hardware is not there yet, with them, if they engage with us, it usually is because that customer has a view that quantum computing hardware will get there, in the same way that machine learning and data analytics did get there and it did impact their enterprises. So these companies, these are very sophisticated Fortune 200 enterprises who are not just for quantum computing but for any emerging technology, are specing or investing small amounts of money to develop expertise in the technology, to figure out how it will fit into their compute workflows and how it might impact their business positively. And so that’s what we do. We’re really providing for them this very technical partnership where we are looking at candidate use cases, so compute bottlenecks they have, and figuring out if and how quantum computing can resolve those compute bottlenecks. For high-value business problems, to allow them to solve them more quickly. And so really, that’s the gist of why a big enterprise today would hire QCWare.

Yuval: Quantum computing has gone through ebbs and flows. On one hand, the quantum computer is on the cover of Time, but on the other hand, there’s talk of quantum winter. Do you see that enterprises are less likely to engage, given what they read in the news?

Matt: Well, that’s a very explicit question. I think that there is enough momentum in the development of quantum computing hardware and technology, to convince these groups that quantum computing will come. But there’s a lot of uncertainty around the timing and around the specific use cases that will be exploitable on quantum computers. And so I don’t see any of these enterprises backing away and saying, “We’ve concluded that there’s nothing there, and so we’re stopping investing.” But they’re moderating investment. And frankly, as they get more information about where quantum computing hardware and algorithms are, that’s helping them have more focus programs. And we’re certainly seeing a lot of that. We’re certainly seeing a coalescing of attention in financial services, for instance, around a couple of use cases, and in drug discovery around a couple of use cases. So we’re certainly seeing that. But that was a very explicit question you asked around quantum winter, and so that would be my answer.

Yuval: How do you differentiate yourself as a company? My sense is that perhaps some years ago, there weren’t that many software companies, a customer could come to you and say, “Hey, let’s run some optimization on D-Wave, and see how it works.” Do you feel like these days you have to develop your unique algorithms or provide value beyond “we can help you get started with quantum computing?”

Matt: That’s precisely correct. I would say there could be up to 200 quantum computing software companies. 200 companies that attest to having some sort of product or product vision around quantum computing. And so if you look at that, those are at… Really, some companies are close to the bare metal and doing this very heavy-duty compilation and translation and error mitigation, and all of these stuff that’s close to the hardware, close to physics. And then there’s a number of middleware companies that are focused on optimizing, mapping or embedding of problems onto hardware, making optimal use of the hardware resources that are there. There’s that level. And then there’s a level on top of that, which is the application layer. And we happen to be one of the players there, I think we’re certainly the largest by a substantial amount as measured by revenue and number of customers, both commercial and government.

So we would differentiate ourselves by, frankly our expertise at inventing and developing quantum algorithms that get baked into application software. In order to develop algorithms, quantum algorithms, it actually takes very deep expertise in quantum algorithm design. And to do that, you need to have a firm grasp of quantum mechanics, of linear algebra. You have to have done quantum algorithm proofs, typically. And so we’ve got a team of close to 50 individuals. And of those, around 40 of them are Ph.D. plus in those fields.

And so our differentiator, the reason people hire us is because they say QCWare seems to be the group that is best at looking at a practical problem that we have, looking at the mathematical structure of that problem, and figuring out how to extend an existing quantum algorithm or to build a new algorithm to allow that problem to be resolved on quantum hardware. So that’s really how we differentiate ourselves. People come to us because they’re trying to answer the question, “How do I get this problem runnable on quantum hardware, this real-world problem?

Yuval: If you look at the spectrum of solution development, you could look at one end at a completely custom job, everything is very specific to every customer. And on the other end of the spectrum, shrink-wrapped software, I go to Google, and I say, “Give me the best directions from Palo Alto to San Francisco.” Where are we in quantum software along that spectrum?

Matt: Right. If you look at completely bespoke versus completely shrink-wrapped or web-based, for quantum software, I would say we’re at step two or three of 10 steps. So we’re way over on the bespoke side still, and that applies to the entire industry. And I think part of that is obviously, there are only a specific subset of problems that appear to have speed-up potential on quantum hardware. I’ll give you an example to eliminate the answer that I have. If you look at let’s say pricing derivatives using Monte Carlo simulation to calibrate pricing models that are used by market makers or traders or investors to price securities, complex derivatives, every one of those firms does it a little bit differently.

So it’s very difficult to have a one size fits all abstraction that would satisfy all of those groups without any kind of customization. And as hardware matures, as more and more banks and insurance companies and market makers get into this, yeah, there will be that kind of standardization. But it’s an astute question that you’ve asked, and the reality is there is still a very heavy amount of this bespoke work being done, at least at the application layer. I think it’s different in middleware and close to the hardware layer, but yeah that’s my view.

Yuval: If so much is custom or bespoke, how do you see the company scaling? So in five or ten years, do you see QCWare becoming a Deloitte? Instead of 50 individuals, you’re going to have 50,000 individuals or 5,000 doing bespoke work, or do you see it moving towards this shrink-wrapped solution?

Matt: Actually, yeah it’s certainly the latter. And we started that rotation about a year ago. So this has all been part of our master plan where we’ve done something remarkable. We’ve signed over 50 customer contracts over the last three and a half years or so with a number of Fortune 50, Fortune 100, Fortune 200 companies. This gave us insight into what products we should build, what software products we should build. And so what we’re doing in the second quarter of this year, we are launching two products that we will be announcing, I can’t announce them yet, but they’re in two domains that lend themselves to being boosted eventually by quantum hardware. These products that we’re launching are going to be on AWS, and they are, in both cases running on high-performance classical hardware. So these are interim products that will allow quantum processing to be slotted into them.

But until that quantum hardware is ready, we’re going to be running on the most powerful classical hardware that’s out there. That’s how we’re making this transition. And in fact… I’m from Minnesota originally, and I and the rest of the leadership team, I think I’ll share the view that we want to have a very, very, very lean team, a very lean and mean team, and we want everyone on our team to be able to demonstrate true ROI. And if you keep that kind of yardstick out there, you’ll avoid in many cases the risk of over-hiring. So oddly, when you think about it, if you have… In a perfect world, it looks like this for a SaaS company, a well-run SaaS company does about $1 million of revenue per full-time employee, a million per employee.

So you could argue that if a company like QCWare with let’s say 45 employees, which is I think at last count what we have, you could argue, “Well, if they’re all focused at building and selling products, you should be able to do 45 million of revenue with that base.” Now, of course, people say, “No. No, you need a lot more customers excited. You got to build this up. You got to overinvest ahead of this and in order…” Maybe that’s true, but I want to see if I can prove and our leadership team can prove that you can run it a lot more leanly than that, that you can scale profitably.

Yuval: Do you have a favorite quantum hardware platform? And do customers come to you and say, “Well, this algorithm is great, but I’d like to see it run on IonQ and then compare it with IBM, and then run it on Quantinuum, and so on and so on? Or do you help them pick one and then run with it?

Matt: We do not have a favorite platform nor do our customers. We are obviously an independent software vendor, and our job is to give our customers the ability to have their problems run… To make them runnable, their production size problems, eventually runnable on anyone’s hardware. So I think it is really fair to say and this isn’t a marketing statement, we have very deep respect for everyone of the teams that are building quantum computers. These are brilliant teams, brilliant individuals, and they’re taking a variety of different approaches. There will be some number that will win and some number that will get consolidated into the leaders, and some that may just not be operating five years from now. But we don’t care to even try to pick a favorite or something, we are just hopeful that all of these hardware companies are going to continue to make advances. And we just want to be for our customers, that software company that is helping them squeeze the most power out of these early generation machines, more power than any other software company can do for them. That’s really what we’re up to.

Yuval: We spoke a lot about customers, and I’m curious if you could give me one or two specific customers that you’re particularly proud of what you are able to do with them.

Matt: Well, first of all, our first couple of customers were Airbus and BMW, and these have been very long-dated engagements for us. And I mentioned those in particular, I love all of our customers. Those two customers have been with us for a very long time, coming up on, I don’t know, four and a half, five years, something like that. And we have deep respect for them. We really love the partnerships. And in both cases, we’ve done a number of different use cases with them building up quantum computing use cases, that should be runnable with potentially significant speedup potential when the hardware gets there. So these problems that we’ve run have been ranged generally from machine learning and optimization. So I’d cite those as being very interesting customers.

The other ones that we can mention are for instance, Roche, and Boehringer Ingelheim. We have a number… I don’t know which ones I can disclose. I can also disclose our investors, and in some cases, they’re also customers, so this will give you some more insight. Goldman Sachs, we can disclose as the customer, they’re also an investor. Citigroup, Koch Industries, Samsung, these are all investors that I’m rattling off. Covestro is a customer and an investor. That’s a smattering. In total, as I mentioned, we’ve signed 50 plus customer contracts, and I think we have at last count something like 32 unique customers. So some of those has been repeats.

Yuval: What do you know today that you didn’t know a year ago, or what has most surprised you over the past year as it relates to quantum?

Matt: A deeper understanding and appreciation of the complexity of building hardware, quantum computing hardware. It is truly like in football, like a battle of inches, it’s an incrementalist process. There tend not to be any particular breakthroughs, it’s just a lot of hard work. I didn’t realize until a year, a year and a half or two years ago, really the extent of that. I just have developed a much deeper appreciation for it.

Yuval: There’s a lot of quantum investment happening in Europe, a lot of government investment. As a US company, do you have a challenge going after some of that money? Do you think that some of it is more limited to EU members, or do you think that the whole world is open to you?

Matt: I think it’s definitely very challenging. We’ve received a couple of very small grants in Europe. I also think it’s probably quite symmetric, in the same way that I think US-domiciled firms have an advantage to go after US government contracts, except basic research where it’s really global. The US government on the basic research level wants to support technology development everywhere. But when you get into applied research and procurement, obviously being US-domiciled will help to get US money. And it’s the same way in the EU. So yeah, it’s very difficult on that side.

On the other hand, although you didn’t ask this explicitly, we have an office in Paris and we are building up extremely strong relationships with our customers in Europe and with various government agencies and nonprofits. So there are rules and practices and policies that we obviously have to and happily abide by, but we recognize as a US company being in other parts of the world, we are not always going to be in the poll position to receive favors from those national governments.

Yuval: And as we come closer to the end of our conversation, I wanted to ask you a hypothetical, if you could have dinner with one of the quantum greats, dead or alive, who would that be?

Matt: One of the quantum greats? Well, I suppose it would be, this is a throwaway common, everyone says this, I guess it would be Richard Feynman. I would like to discuss with him where the field is right now, and whether he, back in 1982 when he published his seminal paper or announced that I think he published in 81 and then briefed it in 82, if things are developing in a way that he thought they would or what… Yeah. So I guess that would be the answer. Yeah. Yeah. And Yuval, I’d invite you to that dinner.

Yuval: Thank you very much. Matt, it was a pleasure. Thank you so much for joining me today.

Matt: It’s been my pleasure. And you have a great day. Thanks very much.

Yuval Boger is an executive working at the intersection of quantum technology and business. Known as the “Superposition Guy” as well as the original “Qubit Guy,” he can be reached on LinkedIn or at this email.

April 3, 2023