Transcript | Quantum Supply Chain Optimisation and Other Use Cases

Supply chain issues are plaguing the world right now across industries. Can quantum computing offer a way to ease the burden and provide real customer advantage? Evidence seems to suggest so. We take a look at this use case and other innovative ones you might not have heard of in this episode of The Post-Quantum World. I’m your host, Konstantinos Karagiannis. I lead Quantum Computing Services at Protiviti, where we’re helping companies prepare for the benefits and threats of this exploding field. I hope you’ll join each episode as we explore the technology and business impacts of this post-quantum era.

Our guest today is the CEO of ZebraKet, Alex Khan. Welcome to the show.

Thank you very much, Konstantinos, for having me on.

I’m excited to have you on. You’re involved with a few startups and, honestly, from what I’ve seen of you, I don’t even know how you’re one person — I’m surprised to see only one person on this call. We joke about many-worlds interpretations sometimes on this show, and here’s an example: You’re literally living many lives all wrapped up in this one world somehow. We’re going to have a great conversation. Why don’t you tell us about how you got started in quantum and this exciting journey?

It was in 2018 when D-Wave made the quantum annealer available in the Leap programme. I happened to see it, and I have always been keeping track of what’s happening with physics and astronomy and VR — any technology that happens to come my way. I signed up for the Leap programme, since it was free, and got into it, got introduced to the quantum annealer, and initially, I thought it was fascinating. I didn’t think I was going to get into quantum computing, but in 2019, I decided to take the MIT expo courses in quantum computing, and that’s where I made a commitment. I actually had a meeting with my family, and I said, “I’m changing directions and going from corporate, the corporate jobs that I’ve had, and I’m going to start working in quantum computing, and I’ve no idea where this is going to lead, but I’m going to commit myself for three years and see how it goes.” I’ve been in it. I’ve been enjoying it, learned a lot and participated in it with many different companies.

That’s amazing. We’re going to talk a little bit more about education later and helping the workforce, and that’s so inspirational: You just made this decision — boom! “I’m pivoting.” And your résumé speaks for itself, and we’ll make it speak out loud in this podcast today. Tell us a little bit about ZebraKet, because we had to pick one title to start with — you’re CEO of this company, so we’ll start there.

Yes, it’s an appropriate example. I was invited to join the CDL programme, the University of Toronto Creative Destruction Lab, which is an incubator that’s becoming very popular, and they are working in many different areas — AI, space, supply chain. I went in there as an individual to see what would happen, and in the interview, I was asked, what is the one thing that I might want to get out of the programme, and I said one word — I said, “Serendipity.” I have no idea what’s going to happen. Hopefully, good things happen.

I met with excellent people — very smart people from all over the world — in the programme. I brought my experience. They were bringing their experience. They were all interested in finding use cases and finding ways to use quantum computing for practical, real-world applications and to start companies. We had many hackathons. We had a project every week, plus we had a hackathon in between somewhere, and in that process, we conceived the idea of doing process optimisation using D-Wave or using the quantum annealer, and that was the start of ZebraKet.

After the boot camp finished, we decided to form the company, and we continued working, and we took on inventory optimisation as the starting point for ZebraKet, but it was in the beginning — we had no idea how we would actually achieve that. And we had mentors — we had Alan Baratz, who was our mentor, and he continued to support us all the way until the end. But we had to show some advantage, and we had to show that our formulation actually had some a benefit in the supply chain industry for inventory optimisation over existing tools that are out there, and that was not easy. We initially had a hard time converting our formulation into something that could be put onto D-Wave. Then we had to find there some advantage, and now we’re in the process of figuring out, can we get some real data and prove that our model actually is useful for practical purposes?

Yes, I was going to ask you about this a little later, but it seems like a good place to jump in now: What can you tell us a little bit about the maturity of quantum optimisation in something like supply chain? I feel like this is a major concern right now in all areas of business. All you hear about is, “These prices,” “supply chain” — everyone’s panicking about it. It’s super interesting that you’re doing work in this area, and what can you tell us? What were you able to accomplish? What kinds of advantage do you think you could bring to this space?

Supply chain right now, with all the disruptions that have happened, is very much in front of us in the news: There were the baby formulary shortages. There was the Evergreen ship that was stuck in the Suez Canal, which disrupted a lot of supply chain. COVID-19 has disrupted supply chains, and initially, it was thought that these supply chains would eventually recover after COVID-19, but when we look at the environment, a lot of contracts have been set up. There’s a lot of lean supply chains that are not set up for any disruptions. Even small disruptions cause a lot of problems; and now we’re having major fluctuations in prices and demand and supplies, and all of that.

I’ve been to a few conferences where the supply chain managers are very much under the gun, very much under pressure, to resolve these issues, and they don’t have the tools to resolve these issues. Inventory managers that probably just focused on their job at one time are now being called in to C-level meetings to explain why there is a shortage or why there is negative publicity, or lack of items for customers. We got into this area because my partner, Ehsan, he’s known about these issues, and he wanted to see what he could do with the optimisation and D-Wave in this area. What we had to figure out was, number one, can we take a practical supply chain problem and embed it on D-Wave?

The first part is, you find the use case — you have to figure out a way to embed it. Now, the techie term for that is to build a QUBO. That’s a quadratic unconstrained binary optimisation model that you have to create. That’s the only way you can put an optimisation problem on D-Wave, or even on gate quantum computers. You have to do that first step, and sometimes, people don’t talk about that.

We’re very interested in what quantum computers can do and what algorithms can be run on quantum computers, but there’s this portion of, how do you take a real problem and set it up so you can actually run it on a quantum computer? We took a few months figuring that out, and we found that it was not any better. The formulation didn’t create any better results than what you can solve with the classical solver, but then we started looking at the components that are removed from these problems.

A lot of these formulations have been simplified over time. They don’t have the quadratic relationships. They don’t have the complexity of the routing problem, packing problem. Any of the difficult problems have been taken out of these formulations to make them run easier on classical computers. When we started putting those components into the formulation, we found that D-Wave did much better — I mean, that’s what it’s built for — than some of the existing classical solvers.

Now, I’m not saying that you can build a genetic algorithm or some specific algorithm that can get to your answer quickly, but when you look at just on the surface, a generic solver that’s in the market, we beat that. In some cases, we had a 40-times-better performance than the industry solver.

It’s not always about full advantage. Sometimes, you could just show that maybe one little piece can be done better in the whole chain and go out from there. When you went through this process, did you start simple, and then try to break the problem? Did you try to just keep adding constraints until it crumbled, or what was your thinking?

I started out with a very core, simple model, and then we started expanding it, making it more complex — bringing in the network multiple retailers, multiple suppliers, thousands of items going through the network, pricing discounts, bundling discounts, customer behavior. Initially, it just started out with a very simple model with one retailer, and it’s become much bigger now.

How do you handle things like bottlenecks? For something like this, you have to get on a queue — you have to wait to use the machine. Is this something that you envision being helpful to run only a few times a day as a result and still show benefit?

At a certain level, you can have optimisation that you just run once a day. If you’re doing some kind of planning and you wanted to optimise a bunch of data, create your optimisation, you get some recommendations and you implement them once a day or once a week or once a month. Those are fair use cases, and you don’t have to use D-Wave all the time for that.

But there are certain instances where if you think about in digital twin, where you have control over your supply chain, you’re getting data about where things are, and then when things shift or the price fluctuates, or somebody calls in and says, “No, I don’t want my shipment,” you want to turn things around very quickly. In an ongoing real-time environment where you’re automating the decision-making process, it is very likely that you want to be optimising on a regular basis and using that information to decide, how do you want to react? Now, you don’t have to totally change your supply chain just because of the recommendation, but you could take components out of that recommendation that have the most cost impact and implement those right away.

In that sense, I see a world, eventually, where companies like D-Wave or gate quantum computers are being sent data where there’s optimisation of some other kind of an algorithm, they are sending back information constantly, and then that information, number one, it’s solved with a lot of data quickly, and then those recommendations can be then used for controlling flights, controlling shipments, controlling various things.

: It sounds like it already meets the criteria for customer advantage as D-Wave introduced it.

Right. The system is available. Yes, there is a queue, but as more funding comes into these companies, they will have more servers and more availability. Large companies, global companies can even eventually decide whether they want to have their own D-Wave system, or other ion-trapped, or something like that can achieve advantage and is sitting in the data center. This is going to happen eventually, and there are companies now that are custom-building quantum computers. There will come a time when companies will see quantum computers as a strategic advantage, and they are not going to rely on sharing resources with somebody else. They’ll have it in their own data center.

Especially for optimisation, annealers are going to be specialised that even when we reach universal gate-based, fault-tolerant, we’ll still have them. That’s going to be very helpful.

Yes, that’s a great answer — thanks. That clarifies that. Before you got started in this, I read about some of your past use cases in quantum optimisation, and that obviously contributed to all of this. What you don’t hear about too often, or maybe ever, is COVID-19 optimisation. I don’t know if you want to explain that, because it sounds like you’re trying to help COVID-19 along — optimising COVID-19. It does a pretty good job on its own.

Yes. That was an interesting project. During COVID-19, D-Wave had a promotion that anyone working on a COVID-19 optimisation use case could have unlimited time on D-Wave, and so there were a few people that reacted to that, and they put some proposals out there. The proposal that I put out there was called Lockdown. It was like a COVID-19 lockdown optimisation — herd versus lockdown. Herd would be, you just let the disease evolve, and deal with it without any lockdowns, and then lockdown would be shutting things down and staying away from the disease.

What I was thinking of, “Between those two extremes, there has to be some optimal, and could we form some kind of formulation where D-Wave could help with that?” I started out with looking at disease evolution. I had to read some papers on the model of disease evolution, or how diseases spread through a population. The model is called SIR — susceptible, infected, recovered. You can add more terms to that like hospitalised, dead and so on.

I started out with that model, and they say it’s a set of partial differential equations. If you have a population that’s susceptible, you introduce a virus into it, and if it’s transmissible or it’s very infectious, it’s going to spread through the population. And there’s a time evolution of how it’s going to move through the population, how many people will get sick and how many will recover and so on.

I’ve built the initial model for that, and then the question was, can we stop the model at some point and see how far the disease has evolved into different cities, and which cities should be shut down to protect the population, while are there some cities where the disease hasn’t moved that far that stay open so you could still have an economy and you can still have a GDP. The balancing act was the limit of people in the hospital, because the whole goal was, you didn’t want to have hospitals fill up, and then you still can’t just shut everything down and not have businesses open and the economy. It was balancing between those two.

A team in India saw my proposal, and they said that they wanted to work with me on that. We started working together. I sent two developers, and we built the whole model and we put a paper out there, which is in the medical archive now, that actually shows that the disease goes through different cities and it moves up in different cities. And then what we did was, in this model, let’s say every 10 days, you evaluate, how far has the disease evolved in these different places?

Then the optimisation, we take the data, we’d give it to the D-Wave to optimise, and then it basically recommends this city should be shut down, while this city should stay open, and it’s trying to maintain some semblance of an economy along with not filling up the hospitals. It’s like a knapsack problem. That’s the hint on how to solve it.

Yes, and it sounds like it doesn’t even matter which strain or variant we’re talking about, because what you’re talking about are the real-world effects. You’re not talking about what the infection rate is initially, or with BA.5, for example, it’s 4.8 times more infectious. That’s not what matters — it’s these other variables that you can gather publicly.

Yes, the team has taken that initial paper and formulation and they’re creating a product out of it and going to the U.S. government’s national health agencies to see if they can use it for any disease management.

I was going to say, it could be a future pandemic tool also. There you go, kids, listening at home — you could adapt the knapsack problem to just about anything. That’s impressive. Yes, I wanted to talk about that. That’s amazing.

We’ve done episodes on portfolio optimisation, and we do those here at Protiviti also for customers. Let’s talk a little bit about what you accomplished back when you wrote your first paper on that, and any thoughts you have on how that’s going forward.

Portfolio optimisation — I met Jeff Cohen with Chicago Quantum at a quantum tech conference in Boston in 2019, and he was talking to Seth Lloyd. I went to that table, and we got to know each other, and then he was looking for, what can he do with the quantum computer? We partnered up, and we were looking for, how can we use a quantum computer for a financial use case, and we tried different ones. We looked at different devices: gate quantum computers like IBM and Rigetti, and D-Wave as well. We landed up on trying to show the financial market that quantum computers can do something. That was our main goal, and, in the end, we decided that we want to visualise the efficient frontier using a quantum computer.

The portfolio-optimisation formulation is a very standard formulation. A lot of people are familiar with it. It’s a simple use case. It is quadratic, because every asset has an interaction with other assets in the market. Some go together, some go opposite or there’s no interaction — they’re independent. That — you have covariances, relationships between the assets can — is a perfect example of a quadratic relationship, and that’s what D-Wave is built for. We took that use case. We had to figure out how to embed it or how to create that QUBO and put it onto D-Wave, and we learned a lot about D-Wave. We learned about the embedding, the annealing time, how to get good results. We were comparing with the genetic algorithm, the simulated algorithm, simulated annealing algorithm.

In the end, after figuring out how we’re going to actually get these optimal results, my goal, as I’ve said, was to draw the efficient frontier and, in the end, we were able to do that. D-Wave, it might not have perfect answers, but it does thousands of samplings. It can find a lot of good portfolios, and it can throw all of those good portfolios onto a plot, and so you can see the efficient frontier. And so we looked at the classical calculations. We looked at D-Wave. We plotted them together, and we found that D-Wave was doing as well as, or sometimes even better than, the classical routines.

Yes, and I urge listeners to quickly Google what efficient frontiers look like, and it’s like a curve of points, and all those portfolios fall somewhere on it, and you’re trying to identify what the best is within the constraints. Where do you see that going in the future? Do you see any projections on what might give us other types of advantage?

A lot of people are looking at financial applications and how they’re going to leverage quantum computers for that. There are a lot of different use cases. The company that I’m working with, QuantFi, they’re using quantum computers for European call options, for example, and they’re using gate quantum computers, which is a very different model. You have to do the modeling differently — different use case, different formulation.

It’s going in two different directions, the way I see it: One is to look for that quantum advantage. That is a lot of work on building a quantum algorithm that has some advantage and finding a use case for creating, or something where you fully leverage the probabilistic nature, the entanglement, the parallelism of a quantum computer, and provide some sophisticated answers. That’s the direction which a lot of people, who say they’re in quantum computing, are typically talking about in building those algorithms.

There is another market that, as you’ve mentioned, is just looking at optimisation and seeing if quantum computers can be used for that. In that case, we are dependent on the hardware. We’ve got Microsoft. Microsoft has a quantum-inspired optimiser, so it can be used for these QUBO problems. It can go to many variables and does a great job. You’ve got Nvidia coming out with their cuQuantum line of products and CUDA and so on, which will simulate a quantum computer, and you can put a variational algorithm or an optimisation algorithm on top of that, and get some result.
And then D-Wave — D-Wave, when we first started, had 2,000 qubits, then, later on, it had 5,000 qubits, then it had the CQM hybrid model on top of that, where last year, we could put 5,000 variables. Now, you can put 50,000 variables on it.

As the hardware improves, as we get better annealers, better gate quantum computers, optimisation will become easier and easier. The algorithms will have to improve. There will come a point where some of these answers can be derived classically, and that’s what we’re all looking for, and we’ll have to get there.

There’s a serious chicken-versus-egg problem: When you start talking about quantum-inspired classical machines that are simulating quantum, may be achieving advantage, and then it’s, like, was it a quantum advantage, or is it like a classical advantage? You’ve mentioned QuantFi. It’s probably a good time to shift gears to that. Do you want to talk about what you guys do over there?

QuantFi is a French quantum startup since 2019. They’ve been in the European market. I met with Kevin Callaghan while he was applying to the Quantum Startup Foundry at the University of Maryland, and I decided to help bring QuantFi to the U.S. market and assist in expanding it in the U.S. market. Initially, for the first year, we were basically in the Quantum Startup Foundry, and the question for all of these startups is, where can you specialise? What can you do that is better than everybody else? Where can you differentiate? Can you create a product that has some value either to the quantum computing community itself or to the industry — banks or healthcare and so on?

It’s hard to give an advantage to existing industry right now, because they have great solvers. They can deal with a lot of complexity already. With QuantFi, there’s a great team that’s working. They’re working with the quantum emulator, financial algorithms, very sophisticated financial algorithms. I decided that one of their products that hadn’t been very visible and wasn’t getting a lot of attention was their quantum emulator, which can emulate 300-plus qubits. It doesn’t take a lot of resources, because it finds ways to compress the quantum circuit where there’s low entanglement, and then it’s able to solve it very quickly.

We have an example where a 300-qubit problem with tens of thousands of gates can be solved in one second on a laptop. Now, it’s not a perfect answer, but you can do a real-world use case using a gate quantum computer, or at least a method of a gate quantum computer, so you’re simulating how your algorithm would work, and you can run it on your laptop. We are wanting to make that emulator available to the market, and so now we’re going forward with that.

And as you know, in the media, there’s been now a lot of discussion about quantum simulators, state-factor simulators, 5,000-qubit simulators, Nvidia using their GPUs to build GPU simulators. We think that we have something unique. We think it can go for a large, complicated quantum circuits and be able to — for someone who’s designing the circuits to be able to actually execute them and learn what the results of those circuits would be.

Yes, there has to be compression of the stuff going, because obviously, you can’t actually represent all the states in 300 perfectly entangled qubits. That would require resources that we don’t have if we converted the universe into a quantum computer, basically.

Right. There are special routines that have to be run for that, but what else would you do? I’ve had conversations with others who are building these complicated algorithms. They have built them, but there’s no way you can run them. The good thing is, we’ve got some focus. The team is very excited. We have a partner, BlueCat, in Japan, that is actually bringing this emulator online. We’ll open it up in the Japanese market. We’re having conversations with a lot of the quantum platform companies to put the emulator in there, and we got accepted into the Duality programme. We’re very excited about that, and the CEO, Kevin Callaghan, is in Chicago in the programme right now.

Yes, I’m excited to see what results you got with that. I’m going to be keeping an eye on that. It’ll be creepy if we get advantage out of that emulator. It’s going to be weird.

Yes, you’re giving up a little bit of accuracy for being able to experiment. That’s the point.

It would be funny if you found some sweet spots, some little piece of a problem that we got something out of. That would be amazing.

Isn’t that what is happening with a lot of the simulated bifurcation machines, the digital annealers?

Yes — like Toshiba.

There will be a period in this NISQ era where these simulators, the quantum-inspired solvers or the annealers, will be the way we solve problems and learn about what quantum computing circuits should look like in the future. We didn’t grow up on quantum algorithms. There will be a whole new generation of students who are going to inherently understand what an entangled circuit should look like, or, if they want to achieve something or come up with some results, what kind of an algorithm they would have to build. I know fluid mechanics and linear algebra and calculus, but this stuff is not something that comes naturally.

There’s no evolutionary reason for our brains to understand this stuff. We’re forcing them to.

That education thing is a good segue here. I did want to ask about what you’re doing to help educate the workforce of tomorrow. I know you actually teach, too, between the hours of 3 and 4 a.m., when you should be sleeping. Let’s hear about that.

I also agreed to help Harrisburg University. This is Terrill Frantz — I met him in one of our Chicago quantum meetings, and he was trying to promote quantum computing in the university, and help out with the ecosystem with education, and he was doing this early in 2019 or 2020. I felt that I had some lessons learned that I could share with students at that time. I volunteered to teach quantum computing, and everything I’ve done is to either help the community or share things that I already know or I’ve learned by experimenting and hopefully, I can help someone to accelerate their journey so they don’t have to spend three years learning it.

I started teaching quantum computing from the basics. We had high school students that we taught during the summer programme — D-Wave, IBM, all the different algorithms: Deutsch-Jozsa, quantum Fourier transform, and then all the way up to Shor’s algorithm, and one of my students that was in that summer camp built a quantum simulator while I was teaching the course.

That was very exciting, and since then, of course, I’ve been doing the undergraduate classes. We set up an undergraduate programme, and we set up a graduate programme. We’re not teaching how to build quantum computers, but what we’ve found is that there’s a huge market of people who are coming to these new startups from various different perspectives — you have marketing people, legal, HR, procurement.

If you just think about everyone — consultants, everyone that’s in a company that goes into quantum computing — or if you have an industry that procures services from a quantum computing company, you have to be able to talk about quantum computers. You have to be able to say, “What is a quantum computer, or how do I evaluate a quantum computer?” If I’m trying to evaluate five companies — I’ve got a quantum product in finance — how would you know which one is better than the other if you don’t even know what quantum computing is?

Beyond the physicists and beyond the people who are making quantum computers, there’s going to have to be a whole workforce that actually understands what this technology is so they can talk about it in their operations meeting or their procurement meeting or in their marketing meeting. As I started teaching graduate-level classes in quantum computing, it wasn’t about, “How are we going to build a quantum computer?” It was educating people who are already in the industry. They were project managers, people who were in various roles in consulting companies, manufacturing companies, pharmaceutical companies who, in their current life, they are not dealing with quantum computers, but you can see very quickly that somebody could pull them in and say, “I’ve got this company talking about solving chemistry problems using a quantum computer. What is that?”

As more people get to know this within companies, they will be able to add value and help those companies to decide, is this something useful for them, or not? I felt privileged. It was exciting when we did project. We had a capstone project in the end where the people would say what kind of quantum computing company they would actually form. These students were beginning to grasp the concepts and the potential advantage, and, hopefully, they can take that message back to their companies. When a highly technical company like us is coming to a company in the industry, there will be someone who can have that discussion with us.

That’s terrific. There needs to be more education all across the spectrum, and I am pushing hard right now: I would love to see a good coder programme, a dedicated undergraduate degree, and you’re ready to code when you come out.

Ready to code quantum circuits.

Yes, these jobs will be there — I’m sure of it. They’re already here. I know they are. It’s only going to grow.

Now, a lot of companies are doing hackathons — IBM, Xanadu. Classiq is now doing quantum algorithm hackathons. A lot of people have been working very hard in this area for decades. I can jump in in 2018, and a lot of the groundwork had already been laid. We’re leveraging a lot of hard work that other people have done. From my perspective, I see that there are people who are building those quantum computers, but there are other people, like me, who are helping connect the industry and people who are not in quantum computing right now and shifting their mindset to at least consider quantum computing.

This has been great, but before I let you go, it sounds like you also have a way to help some folks get educated on quantum computing if they can’t be one of your students. You have a book coming out.

I do. Packt Publishing reached out to me and asked if I could write a book on quantum computing. They wanted me to write a book on Amazon Braket. It’s called Experimentation Using Amazon Braket, and it’s intended for people who are in the industry. If you have a systems analyst or business analyst, and they’ve heard about quantum computing, maybe they know what a qubit is, but I wanted to have a way for them to see what quantum computers can do in the easiest way possible.

I have written down all the code. I show the experiments and how you can take some real-world applications and embed them on a quantum computer, see the results. They don’t have to pay too much attention to the code. Usually, what happens is, I used to have architects where we would go into new technology, like predictive analytics or cloud computing or digital, and we would say, “We need to know more about this — we need a presentation next week” and the person has to learn this new technology in one week, and then talk about it to a bunch of executives.

This book is for those people who, if somebody said to an architect, “We’re already on Amazon Web Services, and there’s a new service called Braket — I want to know what Braket can do for us,” hopefully, within 12 days, if they read one chapter per day, they should be able to have a very good idea of what kind of devices are out there, what can quantum computers do, how can quantum computers be used for optimisation, and then, if they get interested and they want to dig deeper, then they can go into the code, then they can go into all kinds of resources that are available, or they can eventually decide if they want to get a PhD and change their career direction and go into quantum computing.

That’s a huge second step after reading a book. That sounds great — twelve days.

It’s happening to a lot of people.

Yes, a nice 12-day process, at least. Listeners, you want to check that out. Alex, thank you. This has been wonderful. I loved picking your brain, and I’m sure we’re going to be talking a lot over the next few weeks offline too.

All right. I appreciate the time.

Now, it’s time for Coherence, the quantum executive summary, where I take a moment to highlight some of the business impacts we discussed today in case things got too nerdy at times. Let’s recap.

Alex Khan switched careers and started focusing on quantum about three years ago. Since then, he’s accomplished quite a lot, leading to fascinating use cases and startups. The company ZebraKet was born when Alex decided he wanted to try and solve inventory optimization with a D-Wave annealer. Supply chain issues are affecting the globe, which is not surprising considering how little room for error there was in existing systems. ZebraKet spent some time trying to encode classical optimisation processes onto quantum but didn’t see any advantage at first. In a counterintuitive approach, they actually added complexity to the simplified classical approaches that had fallen into use and found that quantum on D-Wave performed better — as much as 40 times better than a classical solver in their lab-based use case.

They then started adding constraints like more retailers to their working model for real-world results. Supply chain optimisation can show benefit when run once a day, which avoids the bottlenecks and long queue times of trying to access a quantum computer more frequently.

Quantum algorithms and approaches can be quite adaptable. During the pandemic, D-Wave asked for proposals to see if their systems could help the scientific community with understanding COVID-19. Alex and team submitted a use case that may help cities and regions determine when lockdowns may be necessary to protect the population. The goal is to optimize the delicate balance between maintaining the economy without filling up hospitals. This is all possible with a clever application of the well-known quantum knapsack problem.

Portfolio optimsation is something we do for customers here at Protiviti. Alex wrote a paper on this back in 2020, achieving excellent results for that time with up to 60 stocks. His other company, QuantFi, is trying to further the advancement of financial applications with quantum and quantum-inspired algorithms as well as a new type of quantum emulator system. Using some compression techniques, this system can emulate 300 qubits, all running on a laptop. It’s not true simulation, as simulating 300 qubits would require turning the entire universe into a classical computer, but this emulation scheme can allow for the creation of large circuits while sacrificing some accuracy. It’s a pretty clever approach to kicking the tires on a quantum algorithm.

That does it for this episode. Thanks to Alex Khan for joining to discuss ZebraKet and their work with innovative quantum use cases, and thank you for listening. If you enjoyed the show, please subscribe to Protiviti’s The Post-Quantum World and leave a review to help others find us. Be sure to follow me on Twitter and Instagram @KonstantHacker. You’ll find links there to what we’re doing in Quantum Computing Services at Protiviti. You can also DM me questions or suggestions for what you’d like to hear on the show. For more information on our quantum services, check out Protiviti.com, or follow ProtivitiTech on Twitter and LinkedIn. Until next time, be kind, and stay quantum curious.