Podcast with Matt Versaggi, Senior Director of AI and Cognitive Technology at Optum Technology
My guest today is Matt Versaggi, Senior Director of AI and Cognitive Technology at Optum Technology, which is part of United Health Group. Matt and I talk about why a health insurance provider is exploring quantum computing, about building internal support inside a 350,000-people organization, and much more.
THE FULL TRANSCRIPT IS BELOW
Yuval: Hello, Matt, and thanks for joining me today.
Matt: It’s a pleasure to be here and a privilege. Glad to do it.
Yuval: So who are you and what do you do?
Matt: So my name’s Matt Versaggi. I’m a senior director of artificial intelligence and cognitive technology, as well as a distinguished engineer here at Optum Technology, and we’re the technology arm of that big insurance giant, United Healthcare.
Yuval: What’s a healthcare company doing in quantum?
Matt: That’s a good question. The first thing you need to do is separate them from life sciences. They always seem to get lumped together, healthcare and life sciences. They are not the same thing. Life sciences are more along the farmer route. They’re going to be more aggressive in reinvesting their dollars towards real innovation. We’re a payer in the payer space. We’re the technology arm of a big payer, but we’re a much more conservative finance company at its core that happens to be in the healthcare space and do healthcare technology as well. So we’re going to take a very different tack to that. We won’t get our commissions from at the top. Ours is a grassroots operation. A number of passionate individuals with long future vision got together and said, “You know, Optum and UHG need to do this quantum thing. They just don’t know they need to do this quantum thing. And we’re just going to do it the best we can with some of our internal partners. And then when we come across something of interest, we’ll let our senior leadership know.” So that’s kind of how that very conservative healthcare space gets moved on.
Yuval: So you are the fan of, “It’s better to ask for forgiveness than permission.”
Yuval: And so now that you have, you’ve sort of come out, what would you say is the business justification for doing quantum for you guys?
Matt: That’s the question, really. And for us, it’s really a couple things. It’s patents and defensive publications. We have a very strong patent program on the wings of a number of really good ex-IBM-ers who came in. And our core justification is to develop patents and defensive publishing for why we exist. And then there’s other ancillary things that really keep us going, like developing code bases and articles and podcasts and speaking events, and really doing all those things that this org needs to do but doesn’t know it needs to do at the senior levels. And so we get that done at the end of the day, but through a different vehicle than the top-down approach.
Yuval: So I’d love to understand the types of projects that you’re working on, that you have worked on for quantum computing.
Matt: Well, ours will be a manifestation of those that could be in the defensive publishing and patenting space. So we won’t be out there actually putting quantum to the test on real hardware under the IBM or AWS or Honeywell banner any time soon. What we’ll do is we’ll educate our people. And we have educated people on three continents, Ireland, India, and the US, in a way that will allow them to grow in this space and pursue patents within their own individual domains. We have so many of them inside of the organization. So that’s kind of our trajectory at this point. We’ve got some work done in the quantum crypto space, because the security folks need to know the kinds of things that they’re up against when it comes to switching from RSA encryption to quantum-safe algorithms. So we do have some actual projects that we’re pushing forward in there, but it’s not like we’re a pharma or a logistics company or aerospace where they’ve got real optimization problems or molecular problems, tinkering with low-level molecules to create new drugs and whatnot. They have real problems that they’re going to really want to try to find solutions for on actual hardware. We’re not in that space yet. However, we’re an extraordinarily rich space, so in time we’ll be a big player.
Yuval: And you mentioned patents. What kind of patents have you filed? I’m guessing some of them have become sort of public information, so maybe you could talk about them now.
Matt: Not yet, and I can’t, unfortunately. Our legal team will be very upset with me, and I don’t want to make that group angry. But we are in that space. We’re being very aggressive about it. And we have a wide swath of opportunities in which to apply that. And so we as well as many other healthcare companies really ought to be in that area preemptively.
Yuval: And so it sounds like your team is more of a prototyping shop, in the sense that you don’t intend to move these projects into production any time soon. Is that fair?
Matt: Right. However, it’s a prototyping shop on the edge of a long future vision and that’s based on situational awareness. And that means that we have a lot of domains where, when quantum matures enough to be affordable, to be doable… And doability is a big deal. Where the material science of holding subtonic particles matures enough that our constituency can use it, then we’ve got a wide base to work with on many different fronts. And we’re preemptively trying to move there.
You have to think that there’s a consumer base and a producer base, right? So the producers make the tech, like your Honeywells near IBMs and Rigettis and all of that, and the consumers… And that’s a big group. The consumers are going to who uses that tech when it gets mature enough that the consumer base is able to more easily use it. Right now it’s very hard to solve problems with quantum, but it won’t be forever. And so what we’re doing is we’re moving this group forward with whatever justification we can have right now, in lieu of waiting for that producer group to mature that tech. And there’ll be a quiescence at some point in time. Once that hits, we’ll be ready to go. And the ancillary concept is that quantum is not a fast follower space. You can’t be a fast follower. It just won’t work. You’ve got to get in there and move your organization to readiness now, so that when those points of quiescence come, you can take off.
Yuval: Can you give me a sense of how large the quantum team is? And what is the composition in terms of, are they computer scientists or are they biologists or chemists? What makes up the team?
Matt: In terms of breadth, we’ve got folks in Ireland, India, and all across the US. So we intentionally have reached that way with our Optum Tech University organization and our patent program. And so we’re actively growing membership in those areas of software engineers who want to learn quantum.
Now, in terms of constructing a team, like an ideal team, a dream team, so to speak… If I were to pick my favorite archetypes, who would be on there? Of course, you’d have your technical people who have that deep domain knowledge in the healthcare space, or whatever space, aerospace, logistics, you name it. The technical people would be there, but they would have a deep domain knowledge and a passion for that along with the quantum skills, because you’re going to need that to be able to ferret out the kinds of problems that you could solve with quantum right now, because it’s very limited, and you want to be able to go right after those. The other thing is that I’d be coupling those with strong business people, because any new of technology that comes in that the business people need but don’t need know they need it, and the technical people can invent it but they don’t know how to sell it, you need those strong business voices in there that can bridge those gaps and take the value and articulate it to the C-suite.
I’d have two other type of stripe people in there, evangelizers who are very good at spreading the word, the gospel of quantum, so to speak, those folks who can speak to the masses in a way that the masses understand. By masses, I mean those folks who aren’t indigenously, quantum folks who, quote, unquote, “get it,” so to speak, but the business folks that they can speak business too. And then the other archetype that I would have, the fourth, would be education, educators, those folks who are really good at teaching people hard stuff so that they get it in the business sense. So I’d have those four stripe of people on my team.
Yuval: I read someplace that United Healthcare is about 350,000 people. And your team, as large as it may be, is probably just a drop in the bucket. I mean, we’re seeing teams of two. We’re seeing teams of seven. We’re seeing teams of 20. We’re not seeing teams of thousands right now. So how do you leverage your relatively small team to stand up competencies in an organization that’s so large?
Matt: Ah, good point. And the way we do it is we have this organization called Optum Tech University. When you have a constituency, that’s that large, you need a way to educate them at scale. And that’s a pre-constructed vehicle that has that level of reach. We can routinely get people across Ireland and the UK. We can get them across the Philippines if we need them. Our mainstays are in India, Ireland, and all across the US. And they have been wonderful to roll out the red carpet for us, allow us to develop these things on our own off the side of our desk a year and a half in the making for our educational programs in the quantum space. And we can roll them out. So now we’re probably pushing upwards towards 30 people who have gone through this program in the last couple of years, and we’re babes in the woods in this space.
It’s also unique because we have to look at the fact that we’re educating people in the consumer side of the quantum spectrum. Every time you pick up a book, whether it’s Dancing with Qubits or any others, they’re going to be for the producer side, educating people in the technology, opaque mathematics, and material science. Now, the consumer side, when you pick up your phone to use it, you don’t care the tech that’s in it. You just care that it works. And so that’s a very different animal to educate and grow. There’s hardly any materials out there, less than 1%, so we had to grow that on our own. That was a significant investment in manpower. But that does allow us to train software engineers in the software engineering part of quantum to move them forward. And that’s how we did it.
Yuval: What’s your view on internal versus external? Meaning, how much external help outside of your organization are you engaging to build a strong quantum computing foundation?
Matt: Well, there’s the tactical implementation of the educational program, and so every everything’s external there, meaning that the thought leadership that went into that came from a few sources that were leading those programs. One’s a guy from O’Reilly Another one writes a book on quantum machine learning from the German space. But it’s all hands-on. Those are materials we can use. But when it comes to the construction of that for internal use, that’s an internal thing for us.
Where you need to turn to the outside is where you can’t get funding, you can’t get support, or there’s unique concerns that you have to address, like the strategy side of things. And that’s why we partner up with groups like the Quantum Strategy Institute, QSI. They’re focusing on a unique space that is meant to accelerate the adoption of quantum out there. And they’ve got experience in the AI space to do that. They know how that works. And we’ve been embedded in that organization to help us understand how the larger enterprise and industry is adopting this, as well as the verticals within that. So those kinds of organizations are where the business strategists and the C-suite need to partner up with so that they can answer a couple questions, like, is this stuff ready to go? Where’s the threat factors when it comes to security? Where are the exploitive opportunities out there from a business standpoint that we can begin to plan for now? And then, based on many of those answers, how do we need to move our organization forward in the right timeframe to take advantage of that? So that’s how I would look at that.
Yuval: You mentioned quantum machine learning, and a couple weeks ago, I interviewed someone on this podcast that’s doing drug discovery, that’s using quantum computing to find the ideal molecules that could attach to a specific virus target, for instance. So do you envision a hypothetical insurance company using quantum machine learning to figure out mortality rates or risk profiles of people that they want to ensure?
Matt: Well, there’s half of that statement I can speak to. There’s probably another half I can’t. I think it would be irresponsible for me to say, “Hey, we’ll use machine learning or any other technology to figure out who we can insure and who we can’t.” I think that violates a lot of our internal laws and ethics.
But what I can say is that when you look at the ability to use machine learning or quantum machine learning, which is really kind of extension of that, to better do any task that you would use machine learning for, then in time we will probably use it. Now, that time has to come across the backdrop of the technology and the software development platforms and the interfaces, and they all have to improve to the point where the consumer base can actually make use of it. And it’s not there yet. But it could be there shorter than you think. So that’s the game that we’re playing, is like, when is it going to be there? How do we know? And the best we can do is take our best minds and constantly monitor those aspects of the advancement of quantum in all of its forms so that we can gauge when we need to put more investment in, move more people around, that kind of stuff.
Yuval: As we get close to the end of our conversation today, I want to do a quick look back into the past and then a quick peek into the future. So starting with the past, you’ve been doing this for a number of years. If you were to do it all over again, what would you do differently? What would you do the same way?
Matt: So I don’t know that I’d do my much differently, because we are in an organization that is going to be very conservative, very reactionary, that would gladly sacrifice innovation when it comes to if that cause risk towards other business objectives that they deem more important. So I think we took the right strategy, doing it at a grassroots level, under the radar, so that we weren’t subject to the crushing return on investment vehicle that is endemic to the healthcare space when it comes to innovation. And we were fortunate enough to have really good folks in the patent place and really good folks in the education space. So we picked our battles well, and we picked our strategy well, I think. And most healthcare companies are going to be in this space as well. Now, the life sciences folks, that’s a whole different animal. They’re very liberalist of spending, because that’s a real element that they need to leverage.
I think I would be a bit more aggressive in getting more people globally involved. We started off in the US pretty well, and we moved over to, to Ireland and the UK and then into India. And we’re still looking to be very aggressive to make this a global kind of a phenomena. I would just do that a little bit earlier.
Yuval: So looking into the future, starting from the consumer side, if you were advising a friend or a colleague when to get started, how to get started with quantum computing, what would you tell them?
Matt: Well, there’s a couple things that would throw on the table right away. First thing is that quantum computing is an OpEx cost, not a CapEx cost. So it’s an operational cost these days, not a capital. You don’t need to take a big refrigerated quantum computer, stick it in your basement, and play with it. And it’s very inexpensive. So the cost is dropping, and it will continue to drop. That’s the first thing. The second thing is that I would get more younger people involved as they come into the program, to make it a common thing, like this is what we’re doing. And then I would also harken in more of the senior leadership to come in, even if it’s on the periphery, and develop ways to reach them better. And we are doing all of those. I would just do those sooner.
Yuval: And on the producer side, if you were master of the quantum universe and controlling everyone’s work path for the next 18 months, what would you have us hardware and software folks do?
Matt: Well, I think they’re doing what they have to do. So right now you’ve got this gigantic geopolitical conflict going on between the US and China that centers around both artificial intelligence and quantum computing. It’s not even hiding. It’s not even a cold conflict right now. It’s out in the open. And what that’s doing is that’s pushing quantum away from being this academic fancy that it was a few years ago to being a huge strategic front and center thing in your face. Like you see governments and academic institutions… You’re seeing political rhetoric in that space. You’re seeing the players in the field move faster. You’re seeing significant investments under the hood that your common person on the street wouldn’t ordinarily see.
So that space has to do two things, has to solve the material science problem to hold that subatomic particle long enough to do information processing on it, and it’s got to do it at room temperature. And if it doesn’t do it at room temperature, it won’t proliferate as much as it needs to. And it will. We’re seeing aspects of that now. There are already PCs on the market that have quantum accelerators at room temperature, so we’ve seen that already. They’ll just have to grow that space. So that’s the first dirty little secret of quantum computing. They can’t hold a subatomic particle long enough to do proper information processing on it. If they don’t get that fixed… And they are, trust me, because the Chinese are working on this too.
The second thing is that the ability to solve problems in the wild, to map them to actual quantum circuits on board or on a circuit chart that provide the algorithm to solve that problem, is horribly immature, horribly immature. And I don’t care who, what vendor says they’ve got that under control. None of them do. Everyone is bad at it. I mean, you’ve got Grover and Shor and they’ve been able to figure out stuff, but they’re brilliant, brilliant minds. The rest of us aren’t even in the same space. So that’s the second dirty little secret, the ability to take a problem in the wild and adequately map that to a quantum solution on a quantum circuit is very, very, very, very difficult. And those guys need to solve that problem.
So there’s this hardware piece, and then there’s this additional piece that goes on top of that, that is needed for the consumer base to get involved. Otherwise the consumer base has to be more quantum specialist than they want to be, and you’re not going to find them there. They just won’t do it. And so once you make that cap to take those two pieces and make them more accessible to consumer base, then that’s going to mushroom, and they’re going to find that that group is going to eclipse the provider group as well. That’s going to change things.
Yuval: Fantastic. So, Matt, how can people get in touch with you to learn more about your work?
Matt: They can find me on LinkedIn. That’s probably the best place. The educational materials that we’ve constructed, we want to open source those. We want to put them out there for the general public to pick up. We just want to add a few more bits to it, to make it a little more robust, to get into the… We’re picking some forays into quantum machine learning, things like that. But that’s a labor of love, and it takes a little while. But we want to open source that so that everyone has access to the consumer side of educating people in quantum. And so if you want to get in touch with me, glad to help out. Just find me on LinkedIn and we’ll connect.
Yuval: Thank you so much for speaking with me today.
Matt: It’s a pleasure to be here. I hope to come back again.
About “The Qubit Guy’s Podcast”
Hosted by The Qubit Guy (Yuval Boger, our Chief Marketing Officer), the podcast hosts thought leaders in quantum computing to discuss business and technical questions that impact the quantum computing ecosystem. Our guests provide interesting insights about quantum computer software and algorithm, quantum computer hardware, key applications for quantum computing, market studies of the quantum industry and more.
If you would like to suggest a guest for the podcast, please contact us.
Originally published at https://www.classiq.io.