Quantum-Si incorporated (QSI) Q2 2026 Earnings Report, Transcript and Summary

Good day and thank you for standing by. Welcome to the Quantum-Si Second Quarter 2026 Earnings Call. [Operator Instructions] Please be advised that today's conference is being recorded. I'd now like to turn the conference over to Risa Lindsay.

Good afternoon, everyone, and thank you for joining us. Earlier today, Quantum-Si released financial results for the second quarter and 6 months ended June 30, 2026. A copy of the press release is available on the company's website. Joining me today are Jeff Hawkins, our President and Chief Executive Officer, as well as Jeff Keyes, our Chief Financial Officer. Before we begin, I would like to remind you that management will be making certain forward-looking statements within the meaning of the Federal Securities Laws. These statements involve material risks and uncertainties that could cause actual results or events to materially differ from those anticipated. Additional information regarding these risks and uncertainties appears in the section entitled Forward-Looking Statements of our press release. For a more complete list and description of risk factors, please see the company's filings made with the Securities and Exchange Commission. This conference call contains time-sensitive information that is accurate only as of the live broadcast date today, August 13, 2026, at 1:30 p.m. Pacific Time. Except as required by law, the company disclaims any intention or obligation to update or revise any forward-looking statements. During this call, we will also be referring to certain financial measures that are not prepared in accordance with U.S. generally accepted accounting principles, or GAAP. A reconciliation of these non-GAAP financial measures to the most directly comparable GAAP financial measures is included in the press release filed earlier today. With that, let me turn the call over to Jeff Hawkins.

Good afternoon, and thank you for joining us. On today's call, we will provide a business update and review our operating results for the second quarter of 2026. After that, we will open the call for questions. Before diving into specific updates, I want to address at a high level the Proteus roadmap update we announced earlier today. We have updated the Proteus launch timeline from the end of 2026 to the second quarter of 2027. The primary driver of the change is the decision to add an additional integrated instrument design cycle to the program prior to moving to production. In connection with this roadmap review, we have also made program leadership and governance changes to strengthen accountability, improve technical oversight and ensure the Proteus program is managed against clear product readiness gates. Furthermore, we are taking operating expense actions in parallel with this roadmap update that are designed to reduce our cash burn and align our cost structure with the revised timeline while preserving investments in the highest priority Proteus work streams. Our conviction in Proteus remains strong. We have generated sufficient sequencing data from integrated units that we remain confident in the underlying technology. That said, we are taking these actions now because we believe it is more responsible and capital efficient to derisk the platform before production than to move forward before product readiness is fully demonstrated. Focusing on the customer experience at launch is the most important thing we can do to protect the long-term value of the Proteus platform. The remainder of the call will be organized around our 3 corporate priorities for 2026, which are as follows: to deliver Proteus with the capabilities customers need, to prepare the market for Proteus launch and to preserve our financial strength. Our first priority is to deliver Proteus with the capabilities customers need. First, let's start by diving a bit deeper into the instrument development. In 2025, we built and tested Proteus prototypes. In the first half of 2026, we built and began testing integrated units. Those integrated units have generated valuable learnings about both manufacturing processes and system-level performance that prototypes could not fully provide. Based on that testing, we concluded that moving directly from the current integrated unit design to production would carry more execution risk than we are comfortable with taking. We're therefore adding an additional integrated instrument design cycle before moving into production. The purpose of this additional design cycle is to reduce manufacturing risk, improve system repeatability, strengthen product readiness and to ensure a positive customer experience at launch. We are making this decision now because we believe it is more responsible and more capital efficient to derisk the platform before production than to scale prematurely and put the customer experience at risk. Finally, I want to reiterate that we have generated sufficient sequencing data from integrated units that we remain confident in the underlying technology. Turning now to recognizer development, we communicated on our first quarter earnings call that our internal developmental sequencing kit was able to detect 17 amino acids. Our progress in this area and the pace of improvement we are seeing has continued to meet our expectations. Given the revised instrument timeline, we will be able to progress even further in this area before locking the reagent formulation for launch. We believe that with this additional time, we will be in a strong position to deliver Proteus with a detection capability of either 19 or 20 amino acids at launch. We will have greater clarity on this specific reagent configuration when we get closer to launch. Finally, I want to share an important update on our Library Preparation strategy. As a reminder, Library Preparation is the process used to prepare a customer sample for sequencing. Our kit performs 2 key functions, digesting proteins into peptides and attaching a linker to those peptides. That linker enables each peptide to attach to the bottom of the nanowells on our sequencing array where it can be sequenced as a single molecule. Historically, we approached Library Preparation as a general-purpose kit intended to work across the broad range of proteins and applications. One of the important benefits of having our first-generation Platinum system in the market is that we have gained significant real-world insight into how different proteins behave during Library Preparation as well as customer needs across a range of applications. Through our work supporting Platinum customers, we have also developed improvements and add-on capabilities related to Library Preparation, both internally and, in some cases, in collaboration with customers. These learnings have led us to an important strategic shift, evolving from a single generic kit to a suite of application-oriented kits that can be used individually or in combination to optimize sequencing performance for specific customer use cases. We believe this strategy provides 2 key advantages. First, by giving customers more complete off-the-shelf tools tailored to their applications, we can make implementation more efficient and reduce the amount of customer optimization required. Second, this approach allows our development teams to deliver a broad range of capabilities faster and with lower technical risk. Developing kits around more specific application needs is less complex than trying to make 1 generic kit, which performs optimally across every protein type and workflow. We are excited about this strategy and believe it positions us to bring new Library Preparation capabilities to market in connection with the Proteus launch. Our second corporate priority is to prepare the market for Proteus launch. As a reminder, during 2025, we launched a placement program for our first-generation Platinum system. This program allows customers to access the Platinum instrument in their own labs without needing to secure budget for an upfront capital purchase. Through this program, we have gained important market and technical insights that give us increasing confidence in the opportunity for protein sequencing and in the alignment between customer requirements and our target specifications for Proteus. First, we have had the opportunity to work with customers across a broad range of applications spanning academic research and biopharma. One recent example is work we have been doing with a biopharma customer in Europe on the application of protein sequencing to adeno-associated viral vectors, or AAV serotyping. AAVs are commonly used as delivery vehicles in gene therapy, and the presence and relative abundance of different AAV serotypes is important to therapy performance. This application is well suited to protein sequencing because the differences between AAV serotypes often involves only a small number of amino acid differences. With Platinum, we have been able to demonstrate the ability to detect relevant amino acid differences. However, due to the limited sequencing output of Platinum, we did not reach the sensitivity required for routine use in that customer workflow. Importantly, using the Platinum data, we were able to model expected performance on Proteus based on its planned increase in number of nanowells, and that model suggests the customer sensitivity requirement is achievable on Proteus. Based on readily available market data, there are more than 150 biotech and pharmaceutical companies developing gene therapies that use AAVs for delivery, suggesting that this single application could represent an attractive market opportunity for Proteus. A second important learning from our Platinum commercial activities is the potential opportunity in non-human proteomics. Earlier this year, researchers at the U.S. Naval Research Laboratory published data using our single-molecule protein sequencing technology for novel pathogen and toxin detection. Over the past few months, we have been working with multiple customers applying protein sequencing to agricultural applications, pathogen detection and typing, and viral protein surveillance applications. We have identified 2 important features of non-human proteomics that make it especially interesting for Proteus. First, many of these markets are not well served by traditional techniques such as immunoassays or Western blots. Second, in many applications, customers need a technology that does not depend on a defined reference because the pathogen or protein of interest may evolve through small changes including single amino acid differences. We believe these market features align well with the core capabilities we are designing into Proteus. Based on readily available market data and looking specifically at the pathogen research and agricultural research markets where our Platinum work has given us a clearer understanding of customer needs, we believe the non-human proteomics market opportunity for Proteus is substantial at more than $4 billion annually. We believe this is a conservative estimate because it does not include additional segments such as environmental, animal health, industrial or food applications where we believe there may be meaningful future opportunity, but where our understanding of the specific customer needs and fit for protein sequencing is still developing. Finally, I want to provide a brief update on our commercial initiatives to build awareness of Proteus in advance of launch. Our commercial team has been executing well, and we continue to receive very positive feedback on Proteus and the wide range of applications it is designed to address. To date, we have identified and qualified more than 250 unique institutions with stated interest in Proteus and the capabilities we are aiming to deliver. These institutions span multiple market segments, including academic research, biopharma and industrial applications, and they have asked to be updated as new information and data become available. We will continue to provide updates on this important market development initiative as we progress toward the Proteus launch. Our third priority is to preserve our financial strength. As I stated earlier, the operating expense actions we communicated today are intended to reduce cash burn, extend our runway and align our cost structure with the revised Proteus timeline while preserving investments in the highest priority work streams required for launch. We are making these decisions as a matter of fiscal discipline. Our focus is to concentrate resources on the activities that most directly improve Proteus readiness, derisk the platform before production and allow us to fund that work responsibly over the extended timeline. I will now turn the call over to Jeff to review our financial results.

Thanks, Jeff. I'll now review our second quarter and first half financial results, discuss the expected financial impact of the operating expense actions we announced today and then provide an update on our full year outlook. Revenue in the second quarter of 2026 was $344,000 compared to $591,000 in the second quarter of 2025. Revenue for the first 6 months of 2026 was $602,000 compared to $1.4 million in the prior year period. Results continue to reflect the dynamics we have discussed throughout 2026, including limited near-term capital purchasing activity for Platinum, ongoing consumable utilization for our installed base, customer awareness of the anticipated Proteus launch and deliberate commercial decisions designed to position customers for a successful transition to Proteus. Gross profit was $172,000 in the second quarter of 2026, resulting in gross margin of 50%. For the first 6 months of 2026, gross profit was $246,000, resulting in gross margin of 41%. Gross margin continues to be affected by the mix and timing of instrument, consumable and service revenue as well as the commercial choices we are making to support the market readiness for Proteus. Turning to expenses, GAAP total operating expenses for the second quarter of 2026 were $25.8 million compared to $30.5 million in the second quarter of 2025. Adjusted operating expenses were $22.6 million compared to $23.8 million in the prior year quarter. For the first 6 months of 2026, GAAP total operating expenses were $49.9 million and adjusted operating expenses were $43.9 million. We continue to manage expenses with discipline while prioritizing the investment required to complete Proteus development, scale internal testing, support manufacturing readiness and prepare the commercial organization for launch. As part of this disciplined approach, earlier today we announced operating expense actions, including a targeted reduction in force representing approximately 20% of the total company headcount. These actions are designed to better align our expense base with the updated Proteus development plan, reduce cash usage, extend our runway and provide additional flexibility as we execute the remaining product readiness work. In making these decisions, we were deliberate in preserving investment in the highest priority work streams required for the Proteus launch. We expect these actions, once complete, to result in approximately $12 million of annual operating expense savings. Net loss was $23.5 million in the second quarter of 2026 compared to a net loss of $28.8 million in the same period of the prior year. Adjusted EBITDA was negative $21.2 million compared to negative $22.2 million in the prior year quarter. Dividend and interest income was $1.7 million compared to $2.3 million in the prior year quarter, reflecting the rate environment and changes in invested balances. As of June 30, 2026, we had $169.9 million in cash, cash equivalents and investments in marketable securities. Taking into account the reduction in force and other operating expense management actions, we now believe we have sufficient capital to support the updated Proteus plan, execute the key activities required for launch and fund operations into the fourth quarter of 2028. This runway allows us to remain focused on the highest return uses of capital, including platform readiness, customer sample evaluations, manufacturing readiness and commercial launch execution. Our full year 2026 outlook remains focused on 3 financial priorities. Maintaining spending discipline, funding the activities required to deliver Proteus with the capabilities customers need and preserving the balance sheet strength necessary to support launch and adoption. For the full year 2026, we're reiterating guidance of revenue of approximately $1 million, adjusted operating expenses of $98 million or less, and total cash usage of $93 million or less. The key financial takeaway is that we are aligning our capital allocation with the updated Proteus development plan. The revised launch timing allows us to complete the additional design, testing and readiness work needed before production, while the operating expense actions announced today are intended to fund that work in a disciplined manner. We're also continuing to invest in the activities that support customer confidence ahead of the launch, including evidence generation with Platinum, Proteus awareness initiatives, customer sample evaluations and targeted commercial engagement. Taken together, these actions are designed to reduce cash usage, extend our runway and preserve the financial flexibility needed to execute the Proteus launch plan effectively. In summary, we remain focused on using capital efficiently, funding the critical path to Proteus launch, and making the right long-term trade-offs to support the adoption of the Proteus platform. With that, operator, please open the line for questions.

[Operator Instructions] Our first question today will come from Scott Henry with Alliance Global Partners.

I'm going to start with a couple big picture questions. I recognize you probably answered this perhaps in more detail, but I'm just looking for kind of a top-down, higher-level thought. What drove the change? Did it come from customers or, as you were working through it, did you just say, hey, if we make these changes, it'll be that much more of an effective product?

Yes, Scott, it's a good question. It comes from our internal assessment. You know, it's a mix of just the performance we're seeing in terms of repeatability and consistency across the machines we have, thinking about what the manufacturing processes have looked like all the way from the optical module and its sort of yields and success rates into the integrated machines, looking at that and saying what improvements are necessary to have a very high-quality repeatable manufacturing and supply chain, looking at long lead components and planning those out. It's really us thinking through all of those things and saying to have the highest likelihood of success to ensure that this is the machine that customers expect, performs the way we want and we can manufacture it and deliver it consistently. We believe we needed to add this additional cycle. It wasn't driven by some sort of new feedback from customers that caused us to believe we need to add some capabilities we weren't planning for. It's really an internally driven assessment of where we are and the best way to get to the product with the capabilities and manufacturing quality we expect.

Okay, great. That's helpful. And with regards to the target of 2Q '27, do you feel pretty comfortable in that number? How much risk is there to that date, I should say?

Yes, Scott, it's a good question. It's a complex development program, so you're never 100% sure of anything. That said, I can tell you that we completed a very thorough review of the program. We took all the learnings from both prototypes and integrated unit testing into consideration. We've added this design cycle that really helps us to retire a lot of key risk before we move into production. And then we've -- as we mentioned in our prepared remarks, really stepped up some sort of the program management side, the things we're doing to have even higher oversight and governance, including from my role in this project. So I think when you take all those together, based on everything we know, we think we've factored that in, that the timeline of Q2 reflects the work required to deliver the product that meets our standards. And that's sort of how we landed on that date.

Okay, great. And then, another big picture question. I think everyone understands the enormity of the proteomics market in the different applications and how each one individual can be a blockbuster indication. But the question is, your new products, they tend to have something that leads the way, a hook. And do you get a sense of, when you go to launch it, what the hook for the Proteus is going to be? Who are going to be the main initial users, adopters of the technology? Or maybe it'll be a mix of multiple. I just want to get your thought on that early adoption.

Yes, I think, Scott, it will probably be a mix of a couple things. I think naturally with any new product launch, and we saw this even in the very earliest days of Platinum, you're going to have some number of customers often in the academic research space who are going to adopt the new technology and sort of explore its capabilities and its edges of performance. I think there's always going to be some number of those folks. That's obviously not your big sort of user base for the long haul, but they certainly are there in the early days. I think as we look at where are more of those applications or market segments where you can sort of have that initial hook. I sort of would think of it in 2 ways. One is clearly in the academic research environment, our ability to have very high coverage of proteins, be able to address the most studied post-translational modifications. Those types of things fit very well into sort of that translational world. I've got a protein, now I want to study it in a population of people. I think as we look more into biopharma and other industrial applications, as we talked about on our call a bit, our ability to work with proteins that aren't covered by existing technologies, we've seen this a bit in the military applications we've talked about, where we've worked across multiple branches here within U.S. military, we've got active engagements with other militaries outside of the U.S. This sort of ability to apply a technology that doesn't need a defined reference to things like pathogen identification, pathogen surveillance, other epidemiological type of work. We think that opportunity could be very meaningful and isn't well served by competitive methods. So that could be another area where we see sort of a hook. And that spans researchers, it spans government entities, it could span industrial as you think about antibody production and other things. So I think those are maybe the 2 really unique capabilities that we see sort of having some hooks into the different segments.

Okay, great. Just a final -- just quick question on R&D. I can see SG&A contracting with the cost cuts. How should we think about R&D in the next couple of quarters relative to Q2?

I think you're going to see R&D fairly consistent to prior quarters. I mean, there are some adjustments as we streamlined the development process as we commented on earlier, Scott. But by and far and large, the spend on the program and the important spend on the program will remain intact because that's the most important thing that we're doing here. There could be some ups and downs just over the next couple of quarters as we change around a few things to make sure the program's on track and it stays on track for Q2 2027. But our operating expense action is really hitting all areas across the company and is really associated with the timeline of the program and to make a few things around the company more efficient. So, from that aspect, I think you're going to see the spend relatively consistent with a few adjustments.

Our next question comes from Michael King with Rodman & Renshaw.

Maybe just a bit of a follow-up on Scott -- your answers to Scott's questions about the additional design cycle. So again, just to be clear on this, was this a function of you running demonstration runs with prototype systems, or was this a result of some scale-up that you were doing for future customer delivery? That's the first part of the question. And then the second part of the question is, how does this affect your ability to satisfy the early access program that you've put in place?

Yes, Michael, so again, this is driven by us having deployed integrated machines in our R&D environment and looking at the repeatability of the performance we see, not just in pure sequencing, as you can imagine when you're developing a product like this, you're also looking at a lot of other more fundamental performance metrics or requirements of specific submodules even of the system. And what we were really looking at is how are all of those things performing? Where are they against what our expectation is? And that feeds back into also, what did we learn when we produced those units and what other things we could do that would make that process more efficient in the future or result in a higher first pass yield. So it's a mix of assessment of the repeatability of performance and also our assessment of opportunities to make some design for manufacturability type changes now. That to -- the second part of your question is critical when you think about when we launch we want to be able to confidently and reliably make these instruments and deliver them to customers. So some of those changes are exactly aligned to do that so we don't end up with that learning down the road when we're in the market.

Okay. And would it apply to the reagents themselves, Jeff, or is this just the instrumentation?

It's very hardware-focused, Michael. I mean, reagent development is still ongoing as well as the consumables, but those areas have really been tracking largely to our expectations. It's very instrument-centric areas on this particular sort of topic.

Got it. And the part of the question about the early access? You slow that down a bit?

Yes, so we haven't yet officially started any early access. We have announced that we have for select customers done some sample testing. With this shift in the timeline, we would expect that the early access will also move out. We don't have an official date with which we're saying we're going to start that activity. But I think from our mindset, we want to really see this instrument design cycle get completed, built and tested internally. That would then allow us to sort of open up for testing customer samples and then sort of following on would be sort of deploying these into the field for early access. So I think that event is going to -- it's going to obviously shift out in time...

Okay. And then is there any thought maybe you could -- would you be able to -- I know you're limited in your supply with Platinum Plus -- sorry Platinum Pro, but are you able to swap those units in just to keep potential clients engaged with the company? Or was that kind of a...

Yes, Michael, we have sufficient Platinum Pro machines to be able to work with customers. We have been continuing to use the placement program where we have folks, as the 1 example I gave, around AAV serotyping. That's a good example of leveraging the placement program to engage with a customer, work on their application, really for us, both the customer and us, to understand exactly what's needed to be there. And we feel very good there about Proteus sort of closing the remaining gap. So we still have that available. Our reps have that available. Some customers like that path, others prefer to wait and start with Proteus. So we really let the customers steer that, but we're comfortable that we have sufficient supply of consumables to continue to support that placement program, as customers may request on the path to Proteus.

Okay. And just as far as kind of merging the full suite of reagents with the commercial unit, assuming that you guys hit your timelines. I was under the impression you'd have all the full repertoire of 21 amino acids available sometime in the first half of next year. Is that still your goal? And would that mean that when Proteus is out there that the full suite of reagents is available?

Yes, so Michael, let's go back. What have we said sort of historically? We've said when we believed we would launch Proteus by the end of this year, we had said we would launch with 18, and we would demonstrate all 20, right, and then add 20 when we got into 2027. What we communicate today was with the timeline move, we expect now that we'll be in a position to launch with either 19 or 20. If we launch with 19, then we would still expect to bring on 20 during 2027. We just don't know yet exactly where we'll land on 19 or 20, but we still think we'll be able to demonstrate all 20 this year. We just aren't yet ready to commit to exactly what will be in that kit except to say we do believe it will be either 19 or 20 given the additional time.

Our next question comes from Kyle Mikson with Canaccord Genuity.

On the additional cycle that you're, I guess, working on, can you maybe speak to how performance or reproducibility could be affected by this? Or maybe the steps that you take to avoid maybe impacting what your prior performance level expectations were going to be?

Yes, Kyle, I think, let me answer what I think you're asking. And if I'm off, just let me know. So what we've observed, again, in the units that we have internally is we have some instruments that really reliably perform at a very high level, meaning above our internal specifications for the product. We have other machines that are performing sort of at or a little below where that's at. And what we are really trying to do with this set of the sort of design spin is make the changes we've identified that we believe allow all of those units to be performing consistently at that higher level. So we think this enables us to sort of unify the repeatability sort of profile of the instruments and also equally as important, take some of the sort of complexity or challenges we saw in manufacturing out by making those design for manufacturability improvements. So I think it will be a combination of the improvement and the simplification of the manufacturing process, but then also the repeatability we should see in the machine should continue to trend up and be consistently above our internal specifications.

Yes, that was perfect. And I guess on this note, I mean just -- I'm curious if the production time or any sort of -- any kind of like timing with -- I don't think the sales cycle is a good question for this, but I think like the production time could be interesting to ask about. Could that be -- is that elongated through this process? It's like not really clear if this is how much is affected in manufacturing versus the kind of the end product, if that makes sense?

Yes, that's a good question. So I'll say a couple of things on the manufacturing front. It has definitely been more challenging and sort of longer to get the machines than we had anticipated. I think we have an optical module that is made as a fully built-up component by one of our partners that is then sort of qualified and shipped to our instrument partner who then puts that into the broader sort of fully integrated machine. We -- when you're doing this for the first time, you certainly learn a lot about how each of the steps in the process works, how bringing all that together works, and then sort of learn a lot about how exactly do you test that and confirm its functioning is going to meet its requirements. So we've definitely learned a lot about that. That improves every time we make another machine, and I think with the set of changes for manufacturability we've identified, we'd expect with this next cycle for that to look a lot more compressed and lead to a very high yield sort of success coming out the end of the line. So I'd say sort of that about the manufacturing side of the house. In parallel, we are obviously being very conscious of how lead times of components can move around. That is not a constraint today, but we're watching very closely, especially in the world of electronics and GPUs, to make sure -- I mean, you don't have to be an expert in AI to know that the proliferation of data centers and the build out there is a massive consumer of those chips. So we are definitely staying very close to our vendor for those and ensuring that we have those procured well in advance so that when we get to the production stage, that's not a limiting factor. So I think that's the most important thing to limit any future production is really staying on top of long lead components. But I think in terms of just the time to build, that should be largely resolved through this next spin we do prior to going into production.

All right, excellent. On the topic of, I think you just referenced data. So with this like chipflation and memory prices increasing and everything, remind us of your exposure to that potential challenge, I guess, beyond '26 when you have Proteus out there? I know the architecture is a bit different, but obviously, like the company historically has been reliant upon semis and everything. So just remind us, please.

Yes, so if you look at Platinum, the consumable is based on semiconductor chip so you're sort of square in the foundry world of producing chips. When you look at Proteus, that consumable is essentially a fused silica array, so form of a glass array that's not a CMOS chip. It has a fabrication step where we put the nanowells on, but in that context there are many vendors that can do that, and we're a very small consumer of that. We also get a lot more individual consumables from a single [ wafer ]. So we -- our exposure to sort of the semiconductor and chip world sort of goes away when we move into Proteus. In terms of the instrument, we still have a little bit of exposure in terms of the GPU. We have a GPU in this system to do data analysis. Again, we're a pretty small player in that world. So not -- we don't perceive it as a large risk. That said, we understand lead times can be very long. And given the relatively low number of these that we use, comparatively, it's very easy to sort of stockpile a bit and hold that to just sort of buffer any potential for that sort of the supply of GPUs to be moving in or out. But we sort of get away from the historical CMOS-related semiconductor chip that was certainly more at risk of the sort of the foundry capacity that you're alluding to.

Okay, that's what I thought. And then on the non-human market opportunity that you talked about, I think that was $4 billion. And I assume it's plant, animal and similar things like that. I mean, I have to imagine that was always part of the plan, I guess. Maybe just kind of backtrack a bit and talk about originally why human, I guess, research, I suppose, was maybe the focus, and especially on the pharma side, of course those drugs are meant for humans, and going forward maybe this is a good like low-hanging fruit potentially for like kind of early on with Proteus and, I guess, with the residual Platinum activities.

Yes, so Kyle, I think obviously we've always wanted to target our technology between markets as we can. I think naturally when you go to market, given all of our respective backgrounds and the backgrounds of our sales professionals, you go to large academic centers, large academic medical centers, biopharma. And when you're in those, you're largely in the human world, whether it's basic research, mechanism of action, translational, you're in, to your point, you're in sort of the human world. Platinum is really showing us really what the breadth of that non-human market might look like. With Platinum, we originally got our first sort of exposure to this through the military side with pathogen and toxin detection. When we opened up the placement program, we got drawn into more, in the area of pathogen ID and surveillance, antimicrobial resistance, lots of different areas sort of getting drawn into different types of laboratories doing that work. Those can be government, they can be industrial, they can be academic. The other side is with the placement program, we got drawn into some in the agricultural space. So really Platinum -- what Platinum did was draw us in, in a way where we got a depth of understanding of what the customer is trying to accomplish, the limitations of the tools available to them, and what specific capabilities we could add that would make our technology even more attractive in that segment. So while it's always been a part of our plan, I think the depth of understanding we've been able to gain from the Platinum exposure really helps us make sure all of those capabilities and the additional that they're looking for are in the Proteus platform so we can really go attack that when we launch the platform.

Great. And last one on AI. So I'm just curious if you're hearing anything, like getting any inbounds or if you're, like, looking forward a bit and trying to think about the potential to generate proteomics data to train an AI model, maybe at, like, a pharma company, biotech company. And just given for clinical human samples, you need kind of more of a -- like there's a wide dynamic range involved there, can Proteus maybe satisfy that market opportunity as well, given that could be relatively larger over time?

Yes, Kyle, I think you're spot on, and I think we're hearing this. Not only are we hearing it, I think other companies operating in the proteomics space are certainly talking a lot about AI. To get the most out of AI, it's really about the training data, how rich is that data? Is it linked to outcomes or phenotypes, those types of things. So we certainly are hearing that in the marketplace. We're hearing it not just in biopharma, but also we hear about it in the academic space because many of the leading academic institutes are sort of operating at the cutting edge of AI tools for whether that be proteomic analysis, whether it be proteins and generating novel proteins or whether it be how might we multi-omic integrate different modes of data and get the most out of that. So I think academic institutes are certainly playing a big role in that area. I'd also say internally we've talked about this before but just to reinforce it, artificial intelligence has been something we have applied significantly in our operations. We use it in the analysis of data. We use it in many other functions. We've talked about it in the recognizer designs, the enzymes in our kit, but we also use it across marketing and finance and other areas, market research inside the company. And it's certainly a very powerful productivity tool as well. So we're bought in. We definitely believe in it as a tool we use, and we do believe our technology will play a role with Proteus in helping people who are looking to build those sort of rich databases to train AI models.

Awesome. I have one more question for Jeff Keyes. On the model, so the $12 million in OpEx savings, yes, like I think it was referenced earlier how that's probably just mostly SG&A, but can you just talk about the cadence and the timing of when that truly will hit and be fully implemented over the next couple of quarters?

Yes, I think we're going to start seeing the benefits of that in, call it, fourth quarter, second half of fourth quarter, because we have to work through severance-related costs and the timeline of certain folks leaving the company. But on an annualized basis, that was my prepared remarks that we expect $12 million related to this specific action of annualized savings as we move forward once it's all implemented.

Our next question comes from Charles Wallace with H.C. Wainwright.

This is Charles on for RK. So I was wondering if you could share kind of the specific gating items between now and the second quarter of '27. And yes, and what new items are in there that weren't in the prior gating items?

Yes, I mean, the main gating item, Charles, to our prepared remarks is adding this additional instrument design cycle. So that cycle involves finalizing the exact changes we're going to make in that cycle, rolling that through manufacturing, building those machines and getting them in-house and tested. So we expect to work through that cycle of finalize the design, build, test and confirm performance over the course of the remainder of this year. So that's really the big task. Assuming that task goes well, we intersect it with what we're doing on the reagent side and with the consumables, and then you bring that together through the first half of 2027 and with the launch in the second quarter. So I would say that that's the key gating item is really just working through that cycle, getting those instruments installed, and confirming the performance.

Yes, that's really helpful. And then I guess another question. So I think you said earlier that you haven't started the early access program, if I heard correctly. And so are you still planning to do that? And when would you kind of do that? Would that be kind of this year event or maybe before the launch?

Yes, so it's definitely before the launch. But yes, I did answer that earlier. You're correct. We haven't started that yet. With this design cycle that we're going through, we wouldn't expect to start that till sometime after this cycle is complete. So, we don't have an exact date, but if you just sort of think through this cycle, confirming the performance, it's probably more of an end of year, beginning of next year type of event is where that would happen. What we're hoping to be able to do perhaps in advance of that is start allowing customers to send some samples for evaluation. Again, we don't have an exact date for when we would offer that capability, but that's something we're keeping a close eye on. We'd like to be able to do that and we would expect to be doing that before we went to a deployed sort of early access. So right now if you're trying to peg something, I'd say, I'd be thinking more as early access is sort of an early 2027 type of event.

Our next question comes from Jason McCarthy with Maxim Group.

This is Michael Okunewitch on the line. Just a couple, primarily on the market opportunities here. First off, I wanted to ask just about the non-human proteomics market, how mature that is and if it's, I guess, ready for a device with the capabilities of Proteus at this time? Or is this something where you would need to build out a base of published academic research first?

Yes, so it's a good question. And what I would say is the reason we focused our remarks on specifically the pathogen research and agricultural is because we've got actual hands-on experience with our existing Platinum and Platinum Pro devices in that market segment. So those are customers trying to solve problems or address initiatives they have right now in place from their institutes. So we believe those markets are absolutely ready for the technology. Some are already applying our first-generation technology and many of them we would expect would move and apply the Proteus technology. Where your comment about the maturity of the markets comes in is the part that we also talked about a little bit in the prepared remarks, which is there are a lot of other areas in non-human like animal health or industrial or food or environmental, those areas, we don't have as much of that direct customer experience. And yet we're doing the work to learn about those. So we just -- we're not really factoring those into our thinking right now, because to your point, we don't know exactly what the fit of the technology is or what the exact need or urgency might be. But certainly, in pathogen and agricultural, given the Platinum work, we think it's a market that's there today and we think Proteus will only build upon the opportunity that we've already sort of uncovered with Platinum.

And then I wanted to see if you could help qualify the difference between only capturing 17 or 18 aminos versus capturing the full suite. Is this something where there's an incremental benefit to the end users? Or does having the full suite open up new applications entirely?

Yes, I would think of it in 2 ways. One is, as you start to get out to 18, 19 and 20, you can imagine that we're talking about, generally speaking, lower abundance amino acids. As you can imagine, when we were developing the technology, we tried to target the most abundant ones. So I think for many applications, there isn't a significant difference between 18, 19 or 20. I think in the area of really deep protein profiling, so maybe somebody who wants to try to sequence as many of the amino acids, let's say, in an antibody as they can. That's where that sort of, to your point, that opportunity opens up where people want to do that, they might really care about the difference between 18, 19, or 20. So I think there are some examples of that. There are many instances where it's probably less impactful to the performance. But I can tell you there's also just a general sort of psychology of customers that I think getting out there to 19 and 20 just sort of resonates as it's complete. It's sequencing. I think when people think of sequencing, they largely think of DNA and they think of, you've got to cover all the bases in DNA. So I think there is sort of maybe a little bit of a mental or psychological bar that it's a less nuanced conversation with the customer when you have all 20 than when you have, say, 18 or 19. But I think outside of a small number of applications where you really need all 20, it doesn't really open up a lot of markets. It's more you overcome that -- it makes the sales cycle a bit more straightforward. It makes the explanation easier. You don't get into all these nuances with a customer about what coverage they need for their application.

Yes, you don't have to justify why 18 isn't enough anymore.

Yes. It's the same concept in the DNA sequencing world. Before we were sequencing everything, we used to use arrays, right? And the question was, well, how many of the different SNPs in the genome could you look at? Well, this one looks at 500,000. This one looks at 800,000. And eventually when sequencing came out, it was like, well, I can just now look at the whole exome or the whole genome, and people just moved there because they didn't need to make that choice. They didn't need to think through exactly what panel they needed if they could just see the whole thing. I think sequencing is somewhat like that in protein. If you just have the full suite, you don't have to really help people think that through despite the fact that in most applications the difference between 18, 19 or 20 is very marginal.

Congrats on the progress.

Thank you.

Thank you. This will conclude today's conference call. Thank you for participating. You may now disconnect.