Quantum-Si incorporated (QSI) Earnings Call Transcript & Summary

Great. Good morning. Welcome to the UBS Precision Medicine Frontiers Summit. My name is Lu Li. I'm the Life Science Tools & Diagnostics analyst at UBS. So happy to kick off our first panel today, which is the New Dimensions in Proteomics and Cellular Research. With me today are Dr. Wenbin Jiang from Cytek Biosciences; and then Jeff Hawkins, CEO from Quantum-Si.

Great. So I guess like the theme today is talking about new dimensions in proteomics and cellular research. So we're thinking about like the new tools, new application in the space. So both of your firms bring in new technology to the space, either like full spectrum flow cytometry or protein sequencing. So I wonder with your new technology, what kind of new application that we're able to do now that we were not able to do in the past?

Sure. I'll start. So Quantum-Si offers what we call next-generation protein sequencing, so reading out individual amino acids. I think what we're seeing customers want to do with our technology, obviously, there's a lot of technologies in the proteomics space, typically fairly dedicated to some sort of set of applications. What people tend to want to do with a new tech like ours is study things that have either been very difficult to study with current technology, maybe requires really expensive equipment or bespoke sort of bioinformatics pipelines or they just couldn't do it all. So single amino acid variant, literally wanting to see if there's a change at one position. They want to look at something called an isomer that they can't detect on mass spec, but they could see it with our tech or they want to look at post-translational modifications, these small changes to individual amino acids that happen when the protein is being expressed. So it's those types of things they want to look at because they believe they're going to be very important in the context of response to therapy or prediction or progression of disease, those types of things. So that's what we're seeing people want to do with the technology. When you're here in the United States, I'd say maybe a little different outside the U.S. where maybe there's less access to proteomics tools in general, and you'll see people wanting to do more basic protein characterization and quantitation. So that's sort of what we see in the 2 different markets.

Sure. And flow cytometer, we are a company, like life science tool company to do flow cytometer. And clearly, it becomes more well known these days because of the BDB Life Science business to [ Waters ] recently. So flow cytometer, look at the name basically to count sales, it's a needle, right? And it looks at the phenotype of the cells and look at the population of different type of immune cells. The conventional technology started quite a few years ago, many years ago, actually. And -- but it has a bottleneck and because of the limitation of the conventional technology with the number of parameters it can look at. So this is where we come from. And we realized the conventional technology actually throw out a lot of information when they detect the cells capture the signals. And so what we come up with what we call the full spectrum profiling technology. Basically, we capture all the information coming out of the detection and from the cells. And from there, that enable us to expand the number of parameters that can detect using a flow cytometer tremendously. And by doing this, that enable many of the applications, which conventional technology wouldn't be able to. For example, one of the example is because it's a spectral, right? So -- and with spectral in the conventional sense, many of the signals, for example, in the cancer cell studies, the autofluorescence from the cancer cell is considered as a noise, which cause problem with the typical detection. But now with our full spectral technology, we treat those autofluorescence from the cell as one of the parameters that enable us to drive the application, improve the sensitivity of the detection substantially that enable us to drive into application, for example, like MRD, right? Of course, MRD typically is one of the application for the conventional flow cytometer, but the sensitivity is very low, a reason why you need to go to a different tool like PCS and sequencing. And -- but with flow cytometer, our technology that improved the sensitivity tremendously to -- by order of -- 2 orders of magnitude. So that can really drive one of the applications, conventional flow cytometry is not being able to. That's one. And second part is in the drug discovery. And as you see more and more new studies, new job being developed, they want speed. And that means they need to look at many different type of parameters. And with our technology, that can really help them to look at lots of parameters very quickly and really speed up the kind of pharmaceuticals early discovery work for the new drug development. Those are 2 typical applications. Of course, there are many due to our technology that have been enabling it.

Great. That's very helpful. I think both of you mentioned that there are like conventional technologies out there and then you bring new one. How do you really drive the new adoptions, right, for your technology? What will be like kind of the key factors out there? And what are kind of like the challenges that you have seen?

I mean, for us, we're truly something new that has never been in the market. So it goes through the challenges, I think, anybody who's ever been a part of bringing a new tech. You have the initial sort of skepticism of does this work at what fidelity with what accuracy, how broadly applicable is it? I think especially in proteomics, people recognize this is not DNA sequencing where sort of you can use PCR and just amplify the low abundant thing and see it. So people understand how hard the problem is. So I think a lot of our early work has been getting the instrument into those leading research centers, whether academic or in pharma biotech, generating the data, proving that it works. So I think a lot of -- academically, it's probably a little more focused on the outcome is the publication. I think with our experience in pharma and biotech is they're often looking to just prove out it's going to work in their systems, whatever their workflow is. They don't necessarily make the data public, but they do all the work to prove it's going to improve their workflow or improve their -- some attribute of -- that they're looking to improve. So for us, it's really that like prove it is a big chunk of it, and that's where we've been focused with the tech.

For us, certainly, we know the market demand, the unmet needs, right? And we found the key pain points. So we start with those key opinion leaders and work with them and develop new applications that really enable them to solve their issues, their problem. And suddenly, they realize, yes, this is a tool I have been looking forward to for a long time, and now it's right there. And clearly, they become so enthusiastic to become your supporter, your endorser. From there, we move actually into one of the key critical applications and for CRO. And clearly, CRO has been working with many pharmas solving their problem, doing work. But once you start to convince the CRO to adopt your technology, of course, you solve their issues, which is to provide solution for almost all pharmaceutical companies with various different needs as well as the costs associated with supporting those customers. If you solve those 2 problems with CRO, and they will jump on to it. This is where we start from as well. And from CRO, they advance us into pharma the other way.

Got it. And we all know that the market has been pretty challenging this year, either from academic side or maybe the biopharma market. So I wonder in this environment, how do you really manage your business? And I think both of the firms are also launching new products this year. So I wonder how do you actually identify the funnel opportunities and go after those very with limited amount of budget?

Yes. I mean, obviously, the academic market has been very challenging this year. I think for us, we're fortunate that we've -- despite being -- we're not a profitable company, but we have -- we just talked about on our earnings call, we have a balance sheet that supports us out into Q2 of '28, which sort of in this market is like sort of almost unheard of for companies at our stage. So I think because of that, we have the sort of the privilege of being able to stay on strategy. So we're obviously very focused on only really investing in the right programs, really managing our expenses, so they're not growing and sort of burning that cash faster, but we don't have to do the big pullback. We're able to stay focused on, okay, as an example, in academia, harder to get capital. Well, we're offering people other ways to acquire the platform. They can rent the platform, they could lease it. We might place it in certain places. So we have some optionality there with a good balance sheet and a fairly low cost to produce our device. Pharma biotech, we haven't seen a big drop off, but it is a longer sales cycle, but we can stay with it and keep working it because we have sort of the balance sheet to support that. And I think R&D-wise, everything we do is really based on people who have been using the tech, what are they saying they want to do next? What are they saying they'd like to be able to do that they can't do, and we just factor that into our pipeline and again, continue to focus on delivering that pipeline, stay on the strategy, don't get distracted by other things, and we're fortunate to have the balance sheet to be able to do that.

Yes. And flow cytometry is a basic life science tool, and you can find it in almost every lab. And so long as they want to do studies in that area, they need a flow cytometer to come along. So there are lots of flow cytometers in the field, and many of them are up for replacement. So we want to -- even though, yes, indeed, the market is tough and the capital expenditure is kind of tight. But as long as they want to do the experiment, they want to work, they need it. And we need to -- so what we do is grab every opportunity with the [indiscernible] and the tools. And clearly, if we can perform with our advanced features, high performance, and we want to outperform the older technology out there to replace them. So that's one thing. And second part is we are global. And clearly, when we see an opportunity in other parts of the world, and clearly, we want to drill down to that. And hopefully, from there, we will be able to capture those business to supplement certain weakness in other territories. From there, that enable us to maintain our business, make it continue to grow, as you can see, even though under today's very tight environment last quarter, and we continue to manage to grow our deployment quantity instrument, our full spectral core technology continue to grow in both number of counts as well as the revenue side.

Great. Since you mentioned the replacement opportunity, I wanted to stick with that concept, right? So I think there are probably around like 50,000 flow cytometers out there. How do you frame your opportunity in that? Like how many can you replace? And what would be kind of like the key factors to really capture those opportunities?

There are 2 aspects, right? One is replacement and the older tools, typically, those tools run 7 to 10 years, and that's the time they need to be replaced. And then we look at them clearly. And when they replace a few aspects. One is the new technology and the new performance. Second part is backward compatibility that need to be insured. Third is I want to make sure it's cost lower. The cost is not only from the instrument acquisition perspective, that also include the maintenance usage as well as the recurring reagent consumption, all those aspects with regarding the cost. That's where we provide them the solution to enable them and to lower the overall cost. Then through Cytek Cloud, which is a very popular platform now to help our users once they get on to Cytek technology, they will be able to reduce their overall cost in designing panel, maintaining their operations. And that's a very important aspect, especially for many pharmas that helps them to reduce the overall operation cost. So that's one. Second part of the business, of course, is to enable earlier, we mentioned about new opportunities, new application, which cannot be done with the old technology that actually helps them to drive towards Cytek, right, because now you are having a tool not only to support your existing needs, but also the future.

And you mentioned -- so we have like some market disruption here given that you mentioned the BD deals, why not go after that. What do you think will be your opportunity with the potential disruption? Any share gain that you can frame about? Any color will be great.

As you can see, we come from the high end of the technology. And clearly, we are leading in this industry. And any disruption from our competition clearly provides us an opportunity because Cytek has already been very well established in this market as a leader. And it's not about competing against with regarding to performance or technology side. It's about whether you can ensure you can continue to support them going for long term and from operation perspective. And even on the recurring revenue side, the reagents, panel design, all those things, Cytek has been working very hard to help them. As you can see, our division revenue has been growing and actually outpaced our instrument growth. That's because leveraging upon our great installed base and our service revenue is growing. This is 2 areas for our business division services plus the instrument continued additional installment that drives our business to grow. And I think anything happened out there is going to help us clearly.

Okay. Great. Maybe lastly on some of the -- I think last time when we talked about, we mentioned -- I think you mentioned some of the clinical opportunity, and you just mentioned MRD will be one of the examples, right? Maybe can you just give us a little bit update in terms of like where Cytek is within the clinical market?

Clinical clearly is depending on geographic location, and you need to go through the clinical clearance country by country. And we started with China first. And in fact, 2 of our tools are clinically approved over there. But not just tool, also including the reagent panels. And so that part, we are well covered. Then from there, we moved into Europe right now. And in fact, our tool has gone through the IVDR clearance, and we have partnered with one of the premier clinical providers over there to drive our clinical instrument adoption in Europe, and we have seen some early traction and the progress also and using our tool to drive the application, earlier I mentioned like leukemia MRD and those kind of new panel design with customers in Europe. And -- of course, coming back, always, we talk about the U.S. and the FDA. And we continue to work through the process right now. It's going to take a while because Europe, China, the clinical approval process is different from the U.S. FDA that we just need to drive this based on the U.S. process.

Jeff, I think for someone that don't really not close to like protein sequencing story, people will always look into, DNA sequencing, which we have seen in Illumina growing, right? I wonder, can we really copy that adoption curve of the DNA sequencing to your story? What would be like kind of like the common threats and opportunities that you have seen so far?

Yes. I think the #1 thing you have to talk about sort of when can you sort of overlay those stories. And what I mean by that is a lot of the team at Quantum-Si came from the DNA sequencing world. In DNA, you have an alphabet of 4 letters and sort of the properties of that are all fairly similar. It's all negatively charged. In protein, the alphabet is 20 letters. There's a tremendous level of sequence context. And without PCR to sort of amplify the low abundance things in proteomics, you have this dynamic range of 10 or 11 logs of dynamic range. So the problem is extraordinarily hard. So I think today, our technology is used in some of these more targeted applications. As it scales over time, it will become a de novo sequencing platform as we bring out our new sort of next-generation platform and continue to improve the level of coverage. So I think there will be a point in time when those will mirror more, but it's quite a bit more difficult to get from 0 to what you might think of in terms of like being able to just drop in a sample and do a whole genome in DNA. I think that day will come for proteomics. I think it's just we're not quite to the beginning of that run. But our belief is when we get to that level of performance, we'll see something very similar to what we saw in DNA, which is when that capability was there and when it was there at a declining cost, it sort of opened up lots and lots of applications and capabilities. So we view it similarly. We just think the journey is a little bit longer to get to that sort of ubiquitous level of capability that we see today in the DNA world.

Got it. Sticking with that point, so what will be the cost point that we're able to kind of unlock the market demand?

I don't know that anyone knows that cost -- proteomics is sort of a fascinating market having spent a lot of time in the DNA world. In the DNA world, you can walk in and talk to a customer and talk about their cost per G or their -- obviously, now some companies wanting to talk more about the total cost of a workflow. In proteomics, it can be all over the board. You could have somebody in a small basic research lab maybe doing western blots and spending $50 or $100. You could have somebody running a big panel of an affinity-based platform that's spending several hundred. You could have someone who owns a $1 million mass spec and then only spends a couple of hundred dollars in reagent. So there's very different -- there's not like one uniform business model in our space, I think, is the more concise way to talk about it. There's some areas that are more reagent heavy and others that are more capital heavy. I don't think right now, pricing is what's holding the market back from sort of that ubiquitous run like what DNA had. I think it's more about the technologies available to do this aren't yet as ubiquitous to do whole proteome or sort of do de novo in the proteome.

Got it. And maybe talk a little bit about your pipeline. Anything that you get very excited in the next 12 to 18 months from your pipeline?

Yes. So we've -- over the last 2 years that I've been here at the company with the management team that's in place now, we've really been sort of iterating and improving upon our technology at the clip of about every 6 to 9 months, improving the sequencing coverage, the sequencing depth, the number of amino acids we can see, also sort of the amount of sample you need. So we have programs across all these different areas. And I think the big program that is slated for a second half of 2026 launch is a platform we call Proteus. So the easiest way to think about it is our current platform, essentially, the optics are in the chip. It's CMOS-based, small benchtop instrument, pretty simple in terms of what the instrument does and sort of the brains are in the chip. The Proteus platform flips that architecture. We go to a very simple consumable. We put the optics in the machine. And just as one example, our current chip has 2 million of these nanowells. So you need more and more wells to take on more and more complex samples, very similar to DNA. The Proteus platform, a chip of the same size, first generation will have 80 million. And it will put us on an architecture that we can scale that through both the consumable size, but also going from point-and-shoot to scanning over time, we'll be able to scale that up into the billions of wells. So this architecture change is key to not only unlocking some opportunities that people would use the tech for today if we had that, but also putting us on a clear path to be able to scale to that de novo level of output that we're going to need. So it's a key program for us. We expect to show data for the end of the year on our prototype machines and then launch in the second half of next year.

Got it. Great. The next question comes like the mandatory. AI is kind of a mandatory question for every panel now. So same for you guys. What's the role of AI in your companies right now? And how important is that? And then how will they evolve in the next few years?

First? Yes, sure. I think there are 2 aspects of the AI. One is on the technology side, how it drives our tool, drives the application. And second part is on the operations side, how AI can help improve our overall efficiency of the company management. We are doing both. And a year ago, we launched one of the software tools to support our image stream and on the data analysis side because AI really is a great tool to help improve the analysis of imaging and simplify the process and make it very efficient and fast. We are continuing to drive this application from there across our old platform on the data analysis side. In fact, one of the -- another tool earlier I mentioned about Cytek Cloud and very popular within the Cytek Cloud, there's a tool called panel design. In fact, it has implemented the AI features. Typically, before it takes about weeks or months to design a panel and to use flow cytometer. Now with our panel design with AI implemented, it can automate the process, make it very fast, pretty much a few hours, you can have a kind of optimized panel out there. So the second part is help to manage the operation overall, and we are working on that. Again, across all our organization, we are looking at using AI to help improve the operation and help to scale up our operation and management.

So I'd agree. I think operationally, if you haven't implemented it in your company, you're probably over-resourced in some areas. Everything from efficiencies of how you run meetings, convert that into meeting minutes into a time line, like these things can be sort of automated through pipelines. Market research departments should be able to be automated into custom GPT. So I think if you haven't done an enterprise-level adoption of AI for business process and business systems and different tools, you're probably overspending on some of those resources. I think the -- when we talk more on the product side, we talk about the tool, but then also when do we have a proprietary set of data to train on. And I think that's an element that not everybody has sort of like completely understood yet at the investor level. And what I mean by that is we've historically used off-the-shelf AI tools to design our recognizers that bind to those amino acids, right? These are engineered proteins. They do not exist in nature. We've used classic tools, computational tools, directed evolution. But what we've done more recently, we just talked about on our call and expect to release some more info. So we have a collaboration with NVIDIA in this area. And what we've done more recently is actually train those AI design tools, but on our proprietary database of over 1 million different candidates where we have inserted mutations, we have information about binding kinetics, about how they sequence, about how pure they are when you make them, how ease -- how stable they are. And we did that, and we did a cycle on that recently. And in one cycle of AI design, we saw a 2x improvement in the coverage of an amino acid that historically, that would have taken many cycles. So we think the tools -- it's not just the tool, it's also when do you have proprietary data to train it on. We use it similarly on the analysis side. So our entire kinetic model is AI sort of derived, and we can take in all the sequencing data happening in the field and internally and retrain that -- sort of regenerate that database on a frequent basis. So we do that a few times a year, and it's a way to sort of continuously improve the product just by learning about all the data that's been generated. So we use it sort of in an external way, but also in an internal way, trained on our database.

Got it. That's very helpful. We only have a few minutes left. Any questions from the audience? Great. If not, then we're going to take a break. And thank you so much for joining me on the panel. And if you guys have any questions, we can certainly do it off stage. Thank you so much.

Thank you.

Thank you.