Kelsey Ortiz, Co-Founder of Photon Queue, on Quantum Innovation

[Music] Welcome back to the Innovations at Research Park Podcast. Student to our show to hear insightful conversations on what innovations research park students, companies, and alumni are creating. My name is Tanmeh Shah and I'm joined in this episode by Kelsey Ortiz. Kelsey is a co-founder and head of design at PhotonQ, a quantum hardware startup developing practical photonic quantum memories to store and synchronize single photons from next-gen quantum communication, computing, and sensing systems. PhotonQ is a highly integrated member of the University of Illinois Entrepreneurship Ecosystem, with investment from Illinois Ventures and physical presence in the enterprise works incubator at the Research Park. PhotonQ is also the first quantum focused startup to spin out of UIUC through a licensing agreement developed with the Office of Technology Management at the University. In this episode, we'll dive into Kelsey's journey leading this startup, her ongoing physics PhD, and talk about a quantum experiment she has onboard the International Space Station. Kelsey, thanks so much for being here. How are you? Good. Thank you so much for having me. This is, I think, one of our first episodes on the show that we will be diving into quantum. I just want to set the stage here about what is quantum, what are quantum memories, what is this overall field all about? Yeah, yeah, it's a great starting question. So basically, what is the field of quantum right now? It's basically doing calculations with quantum particles. I'm most familiar with working with photons, so photons you can think of like a single packet of energy, and they exhibit quantum characteristics. And that means, is unlike regular systems that we have now, you could do two kind of interesting things with them. You can have a quantum particle in a superposition. So if you think of like classical bits in a computer, where you have zeros and ones, a quantum particle can be a superposition of that. So not just zero and one, but a little bit of both. So a little amplitude of zero and a little bit of amplitude of one. That's one interesting thing about the quantum systems. And then another big one is quantum entanglement. So this is when you can have two particles that are entangled. And what they do is they exhibit strong correlations, unlike classical particles. So that would be like, if I kicked one soccer ball, it immediately affects another soccer ball that I have in another room. Right. Classically, that just sounds crazy. But this is what happens with these small particles when you entangle them. So with these two things, superposition entanglement, there's a lot of interesting work that you can do that's kind of going into using that to do calculations faster than classical computers in a quantum computer. And photons aren't the only modality for quantum computers. And there's also atoms that you can put into superpositions and there are different energy states. They're superconducting circuits that you can have in these different energy states. And right now it's kind of like a race of which platform is going to be the winner for creating stable qubits. But my experience is mostly in using photons. So that's kind of like quantum computing using photons or other modalities. But another part of this picture is also the networking aspect of it. So whichever modality wins, right, whether it's superconducting, whether it's trapped ions. For quantum computing, at some point, you're going to want to connect these quantum computers. Or even modularize the computation, kind of like how super computers are done today. And so there needs to be a connection with three and two. And naturally that's always going to be with photons because we call it the flying qubit. And so that's what's interesting to me is not that not only are photons used for quantum computing, but they will be the key to networking these future quantum computers. Very interesting. So I think you sort of answered it, but in the current moment, why is quantum such an interesting space to innovate in? I see there are a lot of startups and companies and even large research organizations that are investing a lot of resources into quantum in recent years. So why is that from your perspective? Yeah, so there's interesting work in trying to figure out ways where quantum algorithms and quantum computers can perform calculations faster than a classical computer. And this will be for specific niche problems. I won't be like replacing your MacBook right now with a quantum computer, right? It would be for specific optimization problems. And this has applications in different sectors, so like automotive, just drug discovery, things like that. So there's a lot of interest in developing these quantum computers to kind of figure out where these niche optimization problems are where it can perform better than classical computers. I would say that other part of it that's interesting is the security aspects of it. So basically quantum computers could provide calculations that break down the security protocols that are used today. And that is a big problem, right? If it does, so it's something that governments are really watching closely right now. Yeah, I think what you're saying is the current best practices or industry standards around cryptography and other encryption methods, those could essentially be deciphered using quantum technology a lot faster than possible with the current state of computers. Exactly. Yeah. And right now quantum computers are developing, so they're not at a stage yet where they're doing that. But I think the just the possibility that they can in the future at a large scale needs to make everybody aware that, okay, now they need to change the practices, figure out different ways to make that more robust. Sure, sure. So Kelsey, what got you interested in quantum? What was your inspiration or motivation to pursue this field of study? Yeah, so I'll go back to where I started. So I did my undergrad at Westchester University in Pennsylvania. That's like a little bit outside of Philadelphia. And was the physics undergrad student trying to figure out where I wanted to go. And essentially I started doing some research in LEDs. So started working with optics and essentially was work to figure out if you can develop an LED that could go to all different colors, like the red, green and blue, just by changing the current. And so that would have impact on how screens are made today and better resolution. So that was kind of my little introduction to optics work and then COVID hit. And I was supposed to be in a lab doing that work for an RU over the summer and basically was not able to do that ended up having to do more of like a software project related to that like coding project. And I realized for me that what I love most is being in the lab and building interesting systems. And so when I was applying for grad school and I got into UI you see, I was looking into the quantum area, but really what caught my eye was Paul Cuyotsk group, which is my research advisor now and kind of looking at the work that he didn't quantum. And it wasn't just doing this abstract physics research like on black holes that nobody, I mean, people care about it, but it's not going to be useful in the next coming years, it was like using entanglement or doing entanglement distribution between two drones. And so that's kind of what drew me in to Paul's group and so joined his group. And I would say another thing that caught my eye about it was he was interested in working in space projects. And for me, that was just kind of a mixture of two things that I didn't really know had an intersection. And it was fascinating to me. So as I got into work on these kind of things. You mentioned Dr. Paul Cuyotk and he's a quite a well known physics professor at the university. As I understand it, the company Photon Q was founded through research from his lab and with fellow PhD students of yours. So what was the origin behind getting it started. Great. Yeah. So basically the work for Photon Q started out. It's been a research project for a couple years now is developing these quantum memories specifically. But it started out with Nathan Arnold who did his PhD work. Yeah, with Paul Cuyotk developing this memory. And I would say my involvement came into play when I was working on seek, which is a compact entanglement source that went to the International Space Station and designed and developed developed it from my first year as a PhD student. I went through that process. It was a grueling process and I can tell you more stories about that. kind of came out of it the other end and it was successful. It was performing well in the space station and that's when I got approached by Nathan with this idea. Basically he had built this quantum memory and presented it at conferences and people were starting to come up to him and say, "Hey look, I actually would like to buy this." And at the time it was just a PhD project for him, so he was like, "I'm not selling this. Like what do you mean? This is just my PhD work." And we also had another co-founder that he was talking to that worked in atomic memories, so a competing memory. And was basically telling him, "Look, the atomic memories are good. They're scientifically very interesting, but they're not as efficient and they wouldn't work as well as these optical memories." And so there is when the idea started brewing of, "Okay, it seems like people want to buy this. It seems like it's competitive in the market." And then he started pulling me into it with my design experience with my project management experience asking me if I wanted to join. And so yeah, I've decided, "Okay, this is something I would love to help you with." And from there, that's where it started. You're the head of design for this company. So what exactly does that entail? Like what are you working on today? What's your core responsibility within the company? Yeah, so just starting out as a founder that basically meant I was wearing a lot of hats. So that means I'm playing a little bit of mechanical engineer on playing a little bit of optical engineer on playing a little bit of electronics engineer to try to build our first prototypes. So Nathan had built this system in an optical lab in a huge optical table that took up basically a whole room. And obviously that's not something feasible if you want to ship this off to a customer. And so when we first started, we were like, "Okay, let's take it one step further. Maybe not jump right off the bat to something super productized." But let's jump into a system that's like two foot by four foot. And it seemed easy enough at first. And so it did the CAD drawings, started figuring out all the subsystems that we would have to be playing with, such as like the optical constraints, the stabilization systems, and then the electronic systems to control it, thinking about user interface. So kind of setting up all the processes to think about and build this system in a deployable manner. And we were able to be lucky enough to have customers that immediately wanted to buy this stuff. And so our first prototype that we were building was we were able to be shipped off to our customers. So we were able to send one to San Dianational Lab and one to University of Maryland. Okay. So two questions there to unpack. What exactly is the product that you are selling or distributing to the customers? And secondly, you mentioned the research lab and the university. Is that the target audience that you're aiming for as a company? Yeah. So first I'll jump into what the core technology is. So ultimately what we would want to do is store photons. Because when you're doing quantum computation with these single photons, a lot of it relies on two photons interfering at a single point in time. And if you have one photon that's received a little bit later, that's interference and things are going to happen. So your computation isn't going to work. And so this is a big synchronization problem. But the issue comes when how do you hold a photon when it's perpetually flying? And if you measure it, it gets destroyed, right? And so you need some kind of quantum memory. And I would say ours is the picks and shovels of quantum memories out there and that it's the simplest, but it works the best. It's the most efficient and can be built right now. It's not necessarily in a deep research and development stage. And how it works is it basically is bouncing off two mirrors. So you can picture photon going into this reflection cell bouncing off two mirrors like a ping pong ball and doing hundreds of reflections. And it creates a beautiful pattern that's actually our photon cue logo. But essentially the photon goes in there. It deterministically reflects back and forth and then exits. And so you're increasing the path lengths of your photon and then storing it for a certain amount of time. And then we do also have configurable delays in which we can switch photons into different loops. So you can have a large delay, a medium delay, and a small delay to have more precision and the time that you want to store your photon. And we do that switching through polarization. So if you have your photon being an h polarization and it goes in between these two PBS's, you can change the polarization and kind of switch it into the loop. And then you change the polarization when you want to pull it out. So now you have this three loop configurable delay. And that's what we developed. Your second question was more looking at what kind of market that we want to look for. So initially we started off with selling products to research labs like San Dia and research groups like UMD. And we want to continue serving those markets. But ultimately we do want to tap into the larger market of the large photon quantum computing companies that are starting to pop up right now. People like cyclonum, like xanadu, like quicks quantum. They're all doing photonic quantum computing. And then I think at a later stage we'll also be wanting to connect with groups like ionq that are doing other kind of modalities because they're also looking at networking their systems. And like I said before, that's going to be happening in photons. So they are going to have a need for these synchronization methods. I see. So you mentioned about a few of these potential customers proactively talking to Nathan and saying, Hey, we want this now. Like this is available on the market. Can you kind of help me understand what is that specific pain point that you're leading with in these customer discovery calls? What is it that they're looking for that photon Q is able to provide? Yeah, that's a good question. So I would say there are two pain points that they're looking for. So the first one would be just large delays. So the first prototype that Nathan demonstrated in the lab was delays of around tens of microseconds with a certain amount of efficiency. Right now, the pain point is that people are looking for delays of around 100 microseconds to be extremely useful, which is difficult to do. But there are things that we were doing to try to figure that out with these reflection cells, so different kind of reflection cells making them larger, but still kind of keeping them compact so that's on a huge massive system. Like we could create a LIGO type of system where you have a huge reflection cell with huge mirrors, but that's going to be extremely expensive, right? So you really have to figure out how to make it as large as possible while keeping footprint down and cost low. And I would say the other pain point in these systems, especially photonic quantum computing systems, is that it depends a lot on single photon sources that are probabilistic. So basically, these single photon sources typically are SPDC-based, so spontaneous, parametric down conversion. And even in the name there, it says spontaneous because it's not happening all the time, right? But you ultimately, when you want to make these computations, you need a single photon source that's coming out at discrete times that you can trust. And so one way to improve this, if let's say you're shooting your photon source and there are times where it's not coming out, is you can reshuffle them using our memories. And so that's where our memories would come into play, but it's a big pain point is trying to figure out good ways to make these sources better in order to actually do these large computations with these photons. I see. So in the next let's say five to ten years, where do you think the quantum computing field is headed? And then that's a broad question, and no one knows for sure, but given your position through photon Q and observing regarding the use of these advanced delay methods and quantum memories and things like that to help accelerate the field of quantum computing in an ideal state in the next let's say decade, what would the quantum computing look like from your lens? So yeah, like I mentioned before right now, it's kind of a competition between different modalities and like which one's going to win and which one is going to produce the best kind of qubits that are stable that aren't lossy. And I think, the quantum computing definitely has a leg up. And certain things, it's kind of leaning towards the silicon way for manufacturing processes already out there. And so it's trying to scale in that way, basically doing a brute force of let's just make this system huge enough that it produces enough photons to do it. And then the other modalities are interesting because they have different benefits like some of them are more stable. So I think if photonic quantum computing wins in that race, photon Q will 100% be necessary to provide the synchronization efforts that are needed to do these computations, right? It'll be necessary in these huge quantum computing warehouses where they have these photonic quantum computers, each one is going to need a synchronization solution. And so I would say in the next five or so years that we're hoping to sell hundreds of these systems out to those companies. And then let's say if another QBIT modality starts to win over, now you jump into the networking area of it. How are you going to now start networking this information? If you want to develop quantum networks across cities, they're going to need these solutions in there. So we hope to not only be in the quantum computing areas, but also in the networking at those different nodes. Got it. How is the relationship been like with research park? I know that you're sitting in enterprise works right now, obviously, and here on the research park campus. But what is that dynamic like? What has it been like to work with folks within research park and interact with them? Yeah, I think it's been a great help. So when we initially started out, the problem was we had a customer and we had to start building these things. But we had no location right where do we start off the lab? And that's kind of where enterprise works came into play, but it was much more than that. I think it gave us connections to people who really just want to help you build your business and gave us access to mentors. People who are recruiters to go out looking for jobs and kind of gives you that community. That's really vital in a startup. Yeah, I think that's one of the key differentiators of the research park atmosphere is that it provides this ecosystem where you can talk to other founders, get advice from them, see the struggles that they've been through and learn lessons from them. It's not easy doing a PhD by itself, but obviously you're doing that on top of working at photon, key and LB lead the company. What has that been like as far as having to do that context which between more academic work and your PhD work and then switching over to business development and design work with photon queue. I'm sure it's not easy. So, you know, what's that experience been like? Yeah, so I would say first I have to be extremely, extremely careful with my time and planning. Eventually I got a feel for, okay, how much is a project going to take. You usually try to add three X times that to be realistic and kind of get an idea of how these things were going to go in the research lab. So I think I've gotten really good at planning and blocking out my time. And I'm pretty strict with it. I'm trying to graduate in December and so every day I'm thinking am I sticking to my goals that is going to get me there right on top of okay, I got to worry about the photon queue stuff as well. But my priority is one of my priorities is finishing the PhD so that I can go full time for photon queue right because also me that's what I want to do. I would say also having good people that work with you. I mean my co-founders are amazing and they've been putting a lot of effort into it and are supportive of me trying to finish this in the time that I have. And yeah, I've been there to help out. Yeah, it's great to have that support network for sure. So I do want to come back to seek. I know we mentioned that a few times now. What exactly is that project and how did you get it onto the ISS? That's a pretty crazy feat. So the seek project stands for the space entanglement in a kneeling quantum experiment. And basically it's a polarization entanglement source. It's SPDC based on a wave guide made by a collaborator that we have at in Montana called 80B or ink. And so one of our goals was to be the first to launch one of these wave guide sources because previously these entanglement sources that have been launched which haven't been many probably like two before us were block optic systems. So it wasn't in a wave guide and you had to basically worry about this light aligning the right way after it launched and experience different vibrations and temperature fluctuations. And so this was a different kind of source that would be inherently more robust to it. So it's fiber picked tail of meaning you don't have to worry about alignment too much. We did have other optical systems after that wave guide system that we did have to worry about alignment. So there was a huge development process there to get that optical system prepped and ready and designed to withstand all these the criteria that it needed to have to launch. So I started this project when I was my in my first year with my PhD started working with Paul Quiat and he basically pitched this project to me and said hey we have six months to do this on a ride share. So it was a cheap option at the time and let's see if we can get this done and I took that opportunity to ran with them was like okay let's do it. I didn't end up not taking six months it took longer than that took three years but I think we learned a lot about launching these kind of systems into space in that time. At the end of the day we launched with a provider called Aegis Aerospace and it went on a SpaceX rocket it flew from Cape Canaveral, Florida in 2024 November 24 and went to the outside of the ISS. So basically he gets on this rocket. Goes to the ISS where astronauts unpack it and then a robotic arm really takes up the package and it puts it on the outside of the ISS and it was important to be on the outside of the ISS because of the secondary goal of the project which was to do a kneeling on the single photon detectors. So the ones that we use are silicon based avalanche photo diodes to measure these single votes on so they have to be really really sensitive right. But that also makes them really vulnerable to other things like radiation in lower orbit and so that's a big problem for these quantum payloads in the future and if you want to make a quantum network in space you're going to have to find solutions for these issues. And so that's something we wanted to probe and look at with this experiment and be able to monitor like how much radiation damage and cats and then how to deal with that and there's two ways we do that. Both through basically heating up the detector but in different ways so we can heat up the detector with a. PC that is a thermal extra cooler that's on the detector itself. Heats it up and takes about an hour to do that and it takes it up to like 80 c and then we have a secondary way where we just shine a bright laser on the detector and it's able to heat it up to about 100 c in like less than a second so it's a much quicker process and so we're looking at the differences in those different processes which ones works better whether we can do laser kneeling and thermo kneeling if that works better. And learning about these methods so deal with this issue. Interesting so why is it a good idea to create these quantum networks in space like what is the benefit of doing that. Yeah so it comes down to if you want to do large distance quantum networking you have two options you can do one in fiber so just set up fibers between cities or wherever you want your quantum network but the fibers can be pretty lossy. And then also you can do free space and you can think of just shooting free space laser or light from one area to another area that you want to go to which would be good if you have a good line of sight but eventually if you want to make these massive quantum networks the curvature of the earth is going to become a problem right and so you're always going to have to eventually go up to nodes in space to be able to reach farther areas for that quantum network and so that's where the motivation came. To start setting them up in space. So do you envision a future where we could have quantum computers running on let's say like satellite nodes in space or like what would be an application of of this research. Yeah so I see more of the distribution of quantum information using these space nodes so I envision a future where we have quantum computers in one city and we have quantum computers in another far away city and we want to connect those two and in that way it would be done via space nodes and you would be distributing the entanglement between those far distances. So it's very cool that sounds like a really exciting prospect for the future. Yeah. Great Kelsey so I want to just wrap up kind of with some words of wisdom you might have for aspiring founders. We've talked about your journey through pursuing a PhD working with an accomplished researcher learning from him and actually taking the step to to co-found a company and actually and get it off the ground. So I think that's some of your key takeaways or any you know words of wisdom that you have to people in similar boats that are looking to start such ventures whether in deep tech or otherwise. I would say first thing is dream big when I first started out as an undergrad in physics I did not think at all that I would a be sending an entanglement source to space and be starting my own company while I'm in grad school. I think never doubt yourself in that way in thinking that oh maybe now it's not the right time I think if you're getting an opportunity and something that really excites you and that you feel passionate about. something you should just follow it. And I think good things will come. Second, I will say is start looking at your problems, day to day in the lab from first principles, which is essentially something that originally, I think was coined by Elan. But I think it really brings true, like start breaking down your problems to the simplest manner and try to think of solutions that way, that way you can iterate fast. I think one thing that I guess bothers me a little bit from academia, I know I'm coming from an academic background is that sometimes things can be slow, like bureaucratic processes, like working with national labs, things are slow and I think they get in the way of you solving your problems, right? So try not to let them do that and really break them down to the smallest idea you can think of and then start thinking about solutions to your issues there. Yeah, and then develop your work that way. - Yeah, I think it's great advice. I really appreciate your time and sharing your insights here today, Kelsey. For our listeners, I hope you really enjoyed this conversation. We'll be sure to have additional topics like this regarding quantum in the future as well. I know that that's been requested. So we will do our best to make that happen. In the meantime, please do check out our other episodes on our catalog and we'll be back with more episodes the future. So take care and we'll see you soon. Awesome. Thank you.