...Hi, welcome to the Canaccord Genuity Global Growth Conference. ` I'm Kyle Mikson, I cover life science tools and diagnostics for Canaccord. Please welcome me to a fireside chat with Quantum-Si. Quantum-Si is the first company to commercialize a protein sequencer platform. We'll, we'll learn more about that today. With the company, we have Jeff Hawkins, CEO. Thanks, Jeff Hawkins, for joining us today.
Thanks for having us.
So yeah, you know, QSI has been public for two years, but maybe the story's not super broadly known. Do you want to just walk through QSI's kind of like mission, strategy, and then technology briefly?
Sure. Yeah, so as Kyle Mikson mentioned, we're the first company to launch a next-generation protein sequencing technology. We launched that in the beginning of 2023, sort of in a controlled launch capacity. Did a lot of retooling of the company that year. We insourced operations, really reset the technology roadmap, from a sort of an innovation perspective, and then just this last quarter, in the Q2 of the year, moved into a full commercial launch of that technology. What it does is, if you think about proteomics, there are technologies that read out protein level. So maybe that's a an array that screens for 2,000 proteins at a time from a sample.
What we do is we can take that same protein, we digest it into smaller pieces, and we actually sequence it one amino acid at a time. So think of it more like the way DNA sequencing works, where you're looking at individual bases, in this case, amino acids.
Okay, all right, that was great. And then, you just had your Q2 update last week. You know, each quarter I feel like it's been, like, a few placements with Platinum, which is your sequencer. And you have an install base of, we estimate, like, just, just over 20, you know, boxes, I guess, out there. Maybe, like, talk about how that has kind of been. I mean, I think obviously there's been, like, an acceleration recently this year with the full commercial launch, but, has that kind of progressed as you would've expected, given it's kind of like a, you know, a novel technology? I mean, you have the, you know, the experience for that.
Sure. Yeah, I think the launch has progressed the way we expected so far this year. As I said, we were a controlled launch last year, putting out sort of a handful of machines each quarter, really sort of white-gloving the service with customers. The last quarter was the full commercial launch, so that was our Q1. You know, still pretty relatively small in terms of commercial, you know, sort of infrastructure, so something we're building to be able to continue to expand. But it's unfolding the way we would expect, you know, a high degree of sort of draw and focus in academic research and starting to bridge over into pharma and biotech, so a pretty natural transition for a brand-new technology. Some customers now starting to begin at major conferences to present data.
So those are good signs that help to build a little bit of the peer-to-peer network, aspect of launching a novel tech like this. And then the innovation roadmap, I know we can probably get into it more here, but, you know, just launched a version 3 of our sequencing kit. We've now made two, improvements to the sequencing kit, you know, in just the last year, so really clicking on the expansion of capabilities as well.
Okay, and then, so a lot of these boxes have been in the field for a year or so, right? Like, I mean, or at least several months or quarters. How has, like, the utilization kind of picked up and recurring orders of consumables and kits from customers?
Sure. So we haven't given out a lot of quantitative information, so maybe talk about it in a few terms. So it in 2023, most of those instruments went into academic research, so think your large research centers like a Northwestern or a Johns Hopkins, Cornell, these sort of where, you know, KOLs are, who do this early work and really publish data. You know, those customers tend to buy, you know, a little more episodically. They'll buy some kits, they'll run their studies, they'll present that data, then they plan their next project and buy. So you'll sort of have a buy and then a lull and then a buy.
As we enter this year, we've talked about starting to get some of our first customers in, pharma and biotech, and those customers, you know, have a bit more of a consistent reorder pattern.
... we're having some interference with your mic, so we're gonna-
No problem. Is this one working? There we go. No problem. So, you know, in the pharma biotech segment, what we're seeing in the early days there is those customers reorder at a much more consistent pace, and that, that to us appears to be linked to just, you know, many more projects happening in parallel within the same lab or same institution, so you see a more consistent sort of order pattern from them.
Got it. Okay, and other applications and use cases that are sort of popular right now, what have you seen, given it's been, like, a year and 1/2 or so?
Yeah, so I would say in the academic research setting, the biggest interest right now seems to be around variant calling and post-translational modifications. So people want to see these very small changes that lead to a difference in the function. So it could be in protein engineering. They're looking at variants to understand when they introduce that variant, what's the response? What's the behavior? It could be in a translational setting, so they want to do a study and look at a given PTM and say, in the disease versus non-disease or the responder non-responder, is the PTM present? If it is, how does it predict the response to therapy or, you know, progression of a disease, that type of thing. In pharma and biotech, we're seeing a lot of traction in something we're calling protein barcoding.
Think of this as attaching a very small peptide, just a small number of amino acids, to the end of it, could be a protein, could be an mRNA therapeutic, and they're able to use this to screen a large number of proteins at one time. They can do this in cellular models, but one of our customers that we talked a little bit about on our call actually did this more recently in vivo in a mouse model, and the luxury for them was they can introduce this, and it doesn't affect the biochemical property of the target, and they're able then to make a measurement and see what was present in that sample after injecting it, in this case, into a mouse.
So barcoding for this sort of screening of libraries of proteins seems to have more traction, or more of the focus in the pharma biotech than in academic, where it's a little bit broader in terms of the types of things they're trying to do.
Okay. So given these learnings, how has that influenced, you know, like new products that you've been launching, kits and so forth? What have you done recently on that front and kind of expect going forward?
Yeah, so I would characterize our technology roadmap in sort of three main work streams. You have the library prep, so that's the actual processing of the sample to get it ready to sequence. You have sequencing chemistry, and then you have essentially the instrumentation and software. So we've been on this consistent sort of innovation cycle of about every 6-9 months for the sequencing kit, so we're on version 3 chemistry now on that kit. What we announced on the call were two new kits coming out on the library prep work stream. So one is a generalized library prep kit to really just expand the types of proteins and the types of mixtures and other things customers can work with, and then a kit very specific to barcoding.
So based on what we've learned from these pharma and biotech customers, really honing in on simplifying that workflow, improving the range of samples they can see, and the sensitivity of that, so that kit will be very dedicated and focused in that area. And then the last one is obviously software and instrumentation. We've had a couple different software releases this year, improving sort of the capabilities, adding other workflows dedicated to certain applications. We haven't announced anything new there yet, but we expect to continue to see those opportunities pop up and then address them with, you know, software workflows that work with those existing chemistries and the equipment in the field.
Okay. And what bottlenecks have you kind of identified, like, as far as, is this cost a big deal? Is like data complexity an issue, and that's why, like, software makes sense to focus on, or yeah?
Yeah, I think in the very beginning of the launch, I think a lot of it was more around just the general awareness of the technology and how it works. Software represented an early opportunity to improve it. The first version of software we ever put out, you know, had, you know, certain limitations and capabilities. So software was a component, but I think now if you were to say, "What do I see as the next most impactful thing?" I think the improvements in the library prep are, you know, the key next thing because it's just unlocks more and more proteins and more and more types of sample types that the customer can work with. In general, our workflow has proven to be very easy for us to implement with customers. Our instrument actually gets installed by the customer.
We don't have a field service engineering team. The instrument gets shipped, they're able to plug that in and run just a sort of a canned little test on the machine that says it's okay. Where we tend to do a little bit of training or support is on the actual wet lab, that library prep and sequencing. But we can go into a lab and, you know, two days later, leave, and there's multiple technicians, scientists, research associates capable of running our technology from start to finish.
Gotcha. Okay, but I got—like, what about, like, proteome coverage and, and, like, amino acids that you're able to, like, kind of detect per run? How—I mean, I thought it's... That kind of funnels into these, like, PTM, kind of like more defined analyses. So how is that being, again, kind of, like, optimized by your work currently?
Sure. So as we launch new versions of the sequencing kit, we're typically addressing a few different things, maybe not all in combination in each launch. But the general themes are, you know, trying to improve the amount of output we're getting in sequencing, and that can come from more effective sort of loading of the chip. It can come from additional recognizers that are detecting the individual amino acids, and it can come from software and the algorithms and being able to retrain those with real sequencing data to be able to, you know, call more and more of those individual amino acids.
So if you looked from version 1 to version 3, the amount of sequencing output a customer would get from a given sample has gone up by more than 5.5-fold, and that's a mix of all three of those things. Recognizer coverage is obviously a big component. As you detect more amino acids, you see more and more within each peptide, but the other factors play into it. With V3, the other improvement that we saw with version 3 is we're now getting even deeper into each individual peptide. So now somebody is getting even more sequence context than they were getting from things they could see before. So these attributes improve each time you improve the chemistry, and it manifests itself as greater output and more visible amino acids.
That then leads to, "I can look for variants. I can look for PTMs. I can look for other events I may not have been able to with a prior version, if we didn't have coverage for that," as an example.
Yeah, so often in like life science tools, the research tool side, and even, I guess, in like diagnostics world, those sort of like analyzers, as you like really just improve and kind of like iterate on chemistry-
Mm-hmm
... and also, like, flow cells and consumables and stuff, you gotta, you have to, like, change your hardware a little bit, your boxes.
Mm.
When does that happen in the roadmap for QSI?
Yeah, right now, all these changes we're describing are compatible with the existing hardware. So someone who purchased this machine back in February of 2023 could be on version three chemistry today, could be using any of these new software or library prep kits. It's all compatible. Right now, our roadmap is all aligned around being compatible with the existing device, which is called Platinum. We did communicate on our last earnings call that we're gonna have an analyst day or investor day coming up at the end of the year, where we'll give a little more visibility into that question, which is, what's the what do we envision as the next generation of sort of hardware and consumables? What would that look like? When does it cut over? Those types of questions we'll, we'll answer at that point.
But right now, all the innovation roadmap we're communicating and we'll deliver is backwards compatible for our customers to that existing instrument.
Got it, okay. Now, taking a step back, talking about the, like commercial model, what is the right way to sort of like, you know, market and sell this type of a product? It's very novel-
Mm-hmm.
Nothing exists like it. You know, the use cases are sort of, some may question them, for example. Talk about like that education, and who's the right kind of sales rep or sales, you know, person, I guess, for this call point?
Yeah, I would say the first thing is I'd say the pro and the con of a novel new tech is everyone wants to talk to you about it, but not everybody is a customer. Not everyone has a budget, not everyone has a budget and a timeframe that's maybe actionable for you as a company. So, you know, one of the key attributes in a sales professional in our world is the ability to sort of engage with that customer, understand their application, their budget, and is this, you know, something that matches the tech today, and they have budget or not, and where to sort of place them in their sort of funnel as they manage that.
So getting someone to take a meeting, someone inbound reaching out to us, people coming by our booth at trade shows, these are things that are a much higher yield today for generating leads than they are when you, you know, are a large company that's been in a market for a long time. I think in terms of the model, we've, you know, erred on the side in the U.S. and Western Europe, the key markets being in a more direct capacity for sales and service, using some channel partners in other areas. We have some partners in Eastern Europe, we have a partner in Japan, you know, continuing to expand that network. But I think we're always looking for, you know, opportunities and partnerships that we could do to further scale.
You know, we'd said on our call, we're adding about 10 more people to our commercial team between now and the end of the year, which is great. That gives us a lot more lift with them than we have today, but we're always on the lookout for other creative ways to just increase that scale and the number of customers we can get in front of every day? 'Cause again, it's not about just the interest people have it, it's how do you really engage and figure out the alignment to their application, our tech, and their budget, to really move it and turn it into revenue.
Gotcha, okay. How does the like strategy differ for these biopharma kind of biotech customers versus the research side? 'Cause now it feels like your kind of your mix now is kind of like skewing more towards biotech a little bit recently.
Yeah, I think we're getting more traction there, but I think your question is really the right one, which is in the academic setting, I think you see two very different flavors of how a customer goes about considering buying our technology. Some folks, on the basis of, you know, presentations we can do, data we can provide, customers they may be able to talk to, you know, will make a decision without having actually personally tried the device. Some people will want to bring the device into their lab and run a demo. So the academic side tends to be a little bit split. Some people will go without a demo, some need it. What we've seen in pharma and biotech is, we're coming in because there's some problem they can't solve with current tech.
They've got resources, they have equipment, they have capabilities, and they haven't solved it yet, and they're looking for us to solve it. So that means they want to demo the product. They want to have the platform and demonstrate that we're gonna solve that problem before they move forward to purchase the equipment. So we, you know, on that side of the sort of the market opportunity, we have seen that demo be a very consistent theme amongst those customers.
Okay, that sounds good. And in terms of like manufacturing the instruments and the consumable products as well, how has that been so far? What, like, you know, how do you even do that? Like, what's your kind of like facilities and whatnot? And then, you know, how does that kind of funnel into, like, your current gross margin on the products?
Sure. So from a manufacturing perspective, the instrument, it's a design and a machine we developed, but we have a contract manufacturing partner who makes that. You know, the team did a great job sort of procuring long lead parts there during some of the, you know, touch-and-go times of COVID and shortly after. So we're in a really good spot there in terms of, you know, key raw materials and the capacity. We can sort of flex it up and down as necessary, so no concerns in that regard in terms of the instrument. When you get to the consumable side, we have the chip, which has a couple of the steps are done outside, and then the remainder are done in-house. Again, a really strong pipeline of materials there.
No concerns about, you know, shortages if we were to get a sudden inflection in sales. And then the reagents. That's one of the things we transitioned in 2023. So that was largely outsourced when I joined the company. We brought the majority of that on the reagent side in-house. And the reason for that is these components are novel. These are engineered by our scientists, the protein recognizers or those enzymes. And that means sometimes they also have, you know, new or novel techniques to produce them or purify them. So doing it in-house, right adjacent to our R&D, is a much more effective way to do that, to get the quality and the consistency we want. But again, very easy to scale. We're doing stuff in fairly small quantities.
Ask anybody who's sort of an expert in protein manufacturing or enzyme manufacturing, and they would tell you, you can scale batches, you know, pretty efficiently. So really no concerns in that regard. In terms of gross margin, to that part of your question, so this last quarter, the gross margin was 57%. Now, because we're still early in commercialization, we do have some components that were acquired in the past, that are at no or low value on the books, so that's adding about 9% in this last quarter. It varies a little bit, but Jeff Keyes, our CFO, on every call, reports out our top-line gross margin and then communicates any of those, you know, sort of one-time variations that might be sort of moving the margin up or down, so people can keep sort of an eye on that steady state.
... Yeah, but again, like, and we've talked about this in the past, how, like, a 50% margin, like a high 40% this early on is pretty good.
It is.
It is, so the semiconductor, but like, you know, chip that you referenced, is that still being manufactured by the Majelac Technologies that you guys acquired a couple of years ago, or is that outside?
Oh, yeah. So, yeah, so Majelac is a company that was one of the vendors. So the way the chip gets made is there's essentially three steps along the way to make the actual chip. The first two steps are done outside by external partners who have, you know, sort of the capital equipment and the infrastructure to do that. Then the chip comes in-house, and that's where the company that used to be a standalone company called Majelac, a company we acquired, we brought that step in-house, and that's the sort of the final packaging of the chip. And then there's a surface chemistry that gets applied that's something our scientist sort of created for our device. So that is done in-house.
So think about that three-step process as a partner, a partner, and then in-house, when it gets to the proprietary sort of steps that are critical to sort of the end performance of it.
Got it. Okay. Well, maybe in the remaining minutes, talk about like the TAM, the market and the excitement, enthusiasm for protein sequencing. So I think, you know, the total TAM could be, you know, over $60 billion or something, you know-
Sure.
Big numbers. For you, what's accessible today seems like maybe 8 billion, 10-
Mm-hmm.
-something like that. So, you know, what, what gives you some confidence or what validates that number, I suppose? And is that, you know, like, why is that realistic? Why should people kind of believe in that, that-
Sure. Yeah, so I think the big number people talk about, whether they use $50 billion, $60 billion, $70 billion, it sort of depends on what you lump into that big number. You know, there's a big chunk of that that is sort of the clinical market or the potential clinical market. You know, we're not playing in that today. So when we talk about what we're working on, we're focused on the research segment, which is about a $20 billion pie. We scale that down to the $8 billion you just mentioned, Kyle Mikson, and that's really us carving the market down for really where we're focused at with our technology today. And as we add new capabilities, then we keep adding to that pie to get up to eventually competing in that entire $20 billion opportunity.
I think when what we're seeing in the field is that $8 billion, things like PTM analysis, things like variant analysis, the protein characterization that is happening at Pharma Biotech, what we're seeing is that's where the traction is. So we, we feel really good that what we're calling our market opportunity is where our technology is fitting and where it's having traction. I think you'll see, and you'll continue to see sort of in our innovation roadmap, where when we're delivering improvements, it's targeted for those segments. And then over time, I think as we bring on, you know, new instrumentation and other things, we'll continue to be able to expand into more and more of the, the next layer. You know, how do we get beyond that into the $20 billion or the $50 billion or the $60 billion over time?
You know, feel really good about the match between those two, sort of the $8 billion and what we've been, you know, effective with, with our technology.
Okay. So that's good to hear. Like, you think it's going to expand, like, beyond the $8 billion in the relatively near term, I guess? Yeah, well, I mean, what unlocks that opportunity for you, and, and, and ultimately, like, is there, you know, does protein sequencing have a chance to be this, like, holy grail of proteomics, basically? You know, the way that next-gen sequencing kind of took over for, like-
Yeah
PCR and so, and things like that.
Yeah, so I came from the genomic space. I was at Illumina for a while. And we have several people on the product development and R&D side from Illumina that have come over to Quantum-Si over the last couple of years. You know, we sometimes joke that it would be so much easier to go back to DNA as hard as that was to solve, and as much innovation that had to occur to make sort of next-gen sequencing in DNA just sort of a household sort of concept. Doing this in proteomics is just significantly more complicated. So we've always been very sort of thoughtful about not claiming we were just gonna take over the world with sequencing overnight, as if it's that easy of a problem to solve.
That said, I think, you know, we are the only people who have been able to deliver a product to the market. I think the innovation roadmap, you know, as I said, we've clicked off from version 1 all the way to version 3, you know, more than fivefold improvement in just 12 months. I think we've hit an inflection point in terms of our innovation. And long- term then, I think to become that ubiquitous tool, you know, the chemistry and roadmap for sequencing and library prep continues, and then the question becomes: how do we address the next level of scale in terms of the output per sample? And that's where the engineering piece, sort of the instrumentation roadmap, sort of comes into play.
How do we go about doing. But that will be the next piece that you add on to ultimately approach, over time, that concept of really de novo sequencing. It's. I think it's doable. I think our scientists would tell you it's doable. I just think we're thoughtful about the roadmap it takes and, and thoughtful about the time it takes to get there. But again, along the way, plenty of unique opportunities to sell the technology that we have today and the various versions we'll do. So we're not worried about having to get to that holy grail to, to be able to generate a lot of meaningful revenue. We think there's a great path to do that, on the trajectory we're on, but we always have that end state in mind, 'cause you have to have that big technology goal to really drive your R&D organization.
Okay. As you think about the technologies that exist today in proteomics, so mass spec and then some of the like, the targeted approaches like ELISA, like, and then I guess, yeah, SomaLogic, Quanterix-
Mm-hmm
... Quanterix and MSD and Luminex and things like that. Where, you know, you guys are kind of unbiased, you know-
Mm-hmm
... in a sense, how should we think about, you know, like, what is—what are you kind of overtaking in of the current kind of, you know, platforms and vendors, sort of? Well, everyone take that, just, yeah, go for it.
Yeah, I think so, you know, for some labs, we're the ability to insource. They've been sending out to a mass spec core lab. They might be spending $1,000- $1,500 for a sample, doing basic protein characterization or identification. They can insource that with us, low capital cost, the level of automation and analysis tools available. But for a lot of customers, we just posted today a new sort of investor deck with a couple of these customer case studies. Oftentimes, we're a complementary technology. We're solving for a portion of proteomic analysis that cannot be done with the existing tech.
So one of the examples we put in there was around being complementary to mass spec for a, for a particular PTM of interest in neurological disease, and the other one was looking at what are called isoforms, which are two things that are different in sequence, but they have the same mass, so you can't see that on mass spec. In this case, it was around calcium deficiency, and we're able to sequence that and show those differences. So there, we're just a tool to take it to the next level because they can't do it with existing tech. So I think, you know, we're really focused on what are the things we can do that our customers just can't unlock today.
Sometimes that's a complement, and sometimes that's a net new capability, but that's really where we focus ourselves, when we're out pitching the tech every day.
Okay. All right, gotcha. So I think with-