Lexicon Pharmaceuticals, Inc. (LXRX) September 9, 2026 Conference Transcript

Welcome to Citi's Biopharma Back to School Summit for 2026. We're very pleased to be introducing our next company, Lexicon Pharmaceuticals. We have Mike Exton, CEO. Welcome. And Craig Granowitz, CMO. And Scott Chianti, the CFO. Welcome, all of you. Thank you so much.

I'm Yigal Nachamavitz. Covered the company for a long time. Well, since before COVID, I think. Anyway, so maybe, Mike, you want to just start us off with the, you know, the quick kind of overview of where things are with, with sodical flows. And of course, you have a very important phase three readout coming up in a critical area of cardiology. So why don't you start with there?

Yeah, no, great. Thanks, Egal. And perhaps before diving into soda, I can give you a little bit of a, you know, where we're at as a company over the last couple of years, really we focused on driving our R&D programs, on establishing financial flexibility and completing a number of partnerships that have gotten us to, I think, a really good state, particularly with soda, but across the board. And now we're on the precipice of a very important data readout in HCM, hypertrophic cardiomyopathy with soda. We expect data in Q1 of next year. And really that is the North Star for the company. And as we sort of execute that, all of the data into the trial, in parallel to that, we start to think around what the commercial opportunity is for us in this pretty important space. And we recently just promoted Rachel Martan in our company to chief commercial officer. Rachel's been leading our BD&L and partnership strategy and was the driver behind our partnership with Nova Nordisk. And I'm really pleased to have her in that position. She ran the insulin businesses for Sanofi and has worked in biotech. And we're going to form a very small but very mighty commercial team with we now have a head of marketing and on the precipice of bringing in a market access person who are really going to do all the pre-launch preparation in HCM because that is our number one commercial opportunity. Beyond that, of course, we're very soon to resubmit Sotogliflozin for T1D. In addition to that, we have pilovapidin, which is a phase three ready program in DPNP. And of course, our program with Novonautisk, which is in phase one LX9851 for obesity. So a lot of things over the next six months where if you think at the end of Q1 of next year, we could be in a situation where we have data from Sonata and HCM, a potential launch in T1D, have started a phase three program in DPNP, and Novo starting the phase two program in weight management for LX9851. So really important time for the company. Lots of things happening, but with a clear focus for us in HCM.

Okay, so let's start with Sonata HCM, as the study is called um now the primary endpoint is this something called this kccq measurement um maybe craig we could just review exactly what that is so people understand um and you're looking for what i understand is like a four to five point improvement there um but you know you've also over enrolled you've exceeded your target so kind of just frame like you know how should we think about having an over-enrolled in terms of probabilities of success and overall you know even without the over-enrollment you know how do you think about the hitting hitting on that that target yeah thank you you got important questions you're asking the primary endpoint that has been

identified by health authorities is improvement in feeling and function since we have a long history with the molecules sodaglifosin and heart failure and not only have improvement in symptoms in heart failure, but also improvement in cardiac mortality and reduction in heart failure events. And outside the US, our partner has gotten approval also for use to reduce stroke and MI with the agent. FDA and we agreed and have, before we started the study, that the primary endpoint would be in this KCCQ score. And the secondary endpoint would be in the New York Heart Association score, change in those scores, placebo-controlled and adjusted. And that is consistent with HCM. The regulatory hurdle is to improve patient functionality and feeling. There are no long-term outcome studies. The CMIs that have been approved do not have long-term outcomes associated with that agent. In part, they needed a dual endpoint because these are unproven agents that have never been on the market. and I think there was a concern that KCCQ score alone would not be enough, and that's why they have a co-primary endpoint, particularly outside the U.S., for a physical function benefit of peak VO2. In our trial design, and I hope as we go to market, we've really tried to take a very broad and pragmatic approach and really think about being first and only in mechanism and complementary in use, no matter what the agent. And the trial design really takes those criteria into account. We have the broadest range of ejection fraction included. We have the broadest range of baseline medications, including CMIs in the trial. And not requiring all of the physical testing like peak VO2 or exercise testing, it also allowed more centers to enroll more patients more rapidly. So that was the philosophy behind the trial. And I think the result of that is we enrolled the trial quickly at relatively modest costs compared to some of these other trials. And as you've already pointed out, we actually over enrolled the study.

Maybe if I just add on to that, Egal, because you mentioned four or five point delta versus placebo. And the study was certainly powered for that. And having an over enrollment improves that powering, obviously. but what our expectation for success is is really being set by the marketplace so a delta placebo adjusted at three points that's statistically significant we think given the benefit risk profile of soda would be a home run for us commercially and anything above and beyond that is obviously upside for us. So we think we have, you know, a great chance of this being a very important part of the armamentarium in HCM.

Okay, so let's just dig a little deeper in terms of the components. So you're obviously enrolling the obstructive and the non-obstructive groups. So now the argument you've made many times is the mechanism is, you know, general to the, you know, cardiac energetics as opposed to being sort of a mechanical MOA where you may see a benefit more in one subset versus the other. So without directly, you know, me answering the question, maybe go into a little more detail there as to how to think about, you know, the non-obstructive versus the obstructive. And then also just when you present the data, you know, how are we going to see it? Are we going to see the groups? Are we going to see everything together? You know, know, just frame that part of it for us. You go, you're doing such a good job answering the question.

I just wanted you to continue. But the trial is designed for the combined group together to be analyzed as one. So all of the 500 plus patients as one primary group. As we've mentioned, the trial is stratified based on baseline obstructive or non-obstructive. But as you've already pointed out so elegantly, the mechanism really is going to be independent of the underlying mechanism. I think there's an expectation in the field, which may or may not be borne out, that the obstructive group is less heterogeneous than the non-obstructive group. The non-obstructive group seems to probably be an amalgamation of different diseases. Certainly, the sarcomeric mutations is going to be a more severe and a relatively more pure form of the disease. It might be more prevalent having these identified underlying genetic mutations primarily in the energy molecules, the energy proteins of the muscle cell, but not uniquely dependent on them. And I think particularly in a non-obstructive group, perhaps one of the reasons why the CMIs have not been as effective on a relative basis is that the disease itself is more heterogeneous. And one of the advantages we believe with soda is its breadth of mechanism. So you're not only getting a number of benefits on the cardiac cell itself with the SGLT1, and as you've already pointed out, you're looking at calcium flux, sodium flux, inflammatory markers, energetics. There's a broad range of activities that have been demonstrated for soda and only soda because there's only SGLT1 receptors on the heart, not SGLT2 receptors. But you're also getting all of the benefits of SGLT2 inhibition on improving the cardiorenal function, which remember we're enrolling only patients with symptomatic disease. So they all have elements of congestion and SGLT2 inhibitors are very good at reducing congestion. So as I've said before, I think the idea is to be first and only in mechanism and complementary in use that we believe that if the profile that Mike laid out bears out, that we would be very high in the use categorization of agents. But at the same time, it's not going to be necessarily exclusive because we have enrolled patients on a CMI in the trial.

Can you speak to the relative split on the populations, or you can't get into that yet in terms of the percent that are HCM versus non-HCM and the percent that are on CMIs that are just based on the general guidelines and knowing the marketplace, what would be expected?

Well, again, we designed the trial pragmatically, and I think with very strong encouragement from our steering committee, We did not cap at a hard number one or the other. I think we enrolled more non-obstructive patients than obstructive. We've been public on that. We have not provided numbers because we haven't finished cleaning all the baseline demographics And we will be presenting that data at an upcoming medical meeting probably in the early part of 2027. But I think we went where the need is. So I think where the centers are right now is there's a greater need and more patients available for clinical trials with non-obstructive disease. So again, we try to model what would be what most people see more closely in clinical practice than having a very rigid clinical trial.

And to the second part of your question around CMIs, we haven't said the exact number. We don't know the exact number yet, but it's relatively small, but significant enough that you can take some data and some questions and some understanding from that. So we've got a cohort, and really this will be the only trial that has a cohort of patients on both a CMI and soda. So it's going to be really fascinating data to see how the two work in tandem. But the other take home from that very point is that patients on a CMI still have the propensity to be symptomatic, which sort of underscores this original point you made, Egal, that the underlying disease truly becomes the energetics of the disease and diastolic dysfunction. And despite improving the obstruction, these patients still remain symptomatic. So there's something much more at play here than just the obstruction. And so we'll see that from this data.

Yeah, I mean, I think, to Mike's point, some of the artificialness of how people think about the disease is a reflection of the trial designs that were done. And if you think about obstruction, it's a continuum. And I think there's this belief just, and again, it's totally understandable, that by running obstructive trials and non-obstructive trials, that they're like these two different things. If you think about a patient, a lot of the issue around obstruction or non-obstruction is how much effort did the person do with their Valsalva. I mean, there's a lot of very qualitative factors. We faced a similar issue when we studied HEFREF and HEFPEF together in our population. So by being able to study the entire continuum, in this case, of outflow tract obstruction and looking at results across a continuum, as opposed to this artificial line, I think it's going to be far more informative to the medical and investor audience about the effectiveness of soda glucosin in this indication.

Plus, you'll also, I just thought about it, plus you'll also get, I know it's, I don't think it's stratified by CMI use, right?

But nonetheless, you're going to get some data on CMI plus soda versus CMI plus placebo that you can put up interestingly and analyze even though that's not stratified it's not you know that's not the main thrust of the study but that's there's some very interesting data there potentially yeah and so to that point presumably every patient that was on a cmi when they were placed on the cmi because they just need to be on a stable dose so that means they've been on a cmi in six months because that was the entry criteria so they were all on map because afi wasn't commercially available in that time frame we didn't stratify or hold them accountable to a certain baseline um outflow tract obstruction so i can tell you there's going to be continuum of outflow tract obstructions some of which would be considered classically obstructive and some would be at the baseline non-obstructive so to mike's point there is not a great correlation between obstruction and symptoms and we have all of those flavors in the trial yeah it's always it's always dangerous when you try to take uh create two categorical variables out of what's a continuous variable for we'd love to do it but uh you're right it's uh it it is fraught with danger and misinterpretation yeah um okay um by the way so we should talk to about some of the secondary endpoints please what are the ones that are most important on in the study well i think the ones and you saw some of that data presented at ESC from the AFI study from Acacia. But there's really beyond the New York Heart, which is basically the provider assessment of patient functionality. So sort of the mirror image, the patient with KCCQ is rating themselves. New York Heart is the healthcare provider sort of making an external assessment to the patient. But in general, they're very highly correlated over time. There's really a lot of parameters. I would put them in the function and feeling So those are the primary and secondary endpoint. And then there's a lot of secondary endpoints, the enzymes, and then really looking at a lot of the echo parameters. And I think the echo parameters are really important. And I think to me, one of the key takeaways from ESC is that a lot of the benefits, I would say, and the speaker on the podium said, these benefits are modest and reversible. And so, you know, that to me is an important element, not just looking at the enzymes, because I think the enzymes are not very reflective, at least with CMIs, on benefit, because the biggest benefit is on cardiac enzymes, but there's no correlation of that to anything. I think looking at things like left atrial volume is going to be really important, because that's a real reflection of left ventricular function. The less functional the left ventricle is, the more you generally get a large left atrium. We're going to look at left wall thickness. We're going to look at the speed of the contraction. So there's a lot of echo parameters that are going to be looked at as either secondary or exploratory endpoints. okay um and i imagine if you if you hit presumably on the first on the primary endpoint that you would expect to see the the secondary endpoints track accordingly if the patients are reflecting a you know better feeling better uh you know quality of life right i have a high degree of confidence in that you go because again we've got this massive database of heart failure outcome studies. And like in HEFPEF, the primary endpoint of the trial, we've talked about that SOTA, Picardia, was left ventricular wall thickness. And that was a primary endpoint and was achieved. And KCCQ score was beneficial and six-minute walk and a number of other parameters. So I have a high degree of confidence just based on our prior experience in heart failure that you're going to see benefit across a range of these parameters.

Okay.

Did you want to say anything else about acacia to read through or any other comments there in terms of thinking about your study i just think they're complementary mechanisms yeah um and the fact is we're going to have patients on a cmi in our trial i think that's going to give people comfort around relative safety and tolerability of that in a way to mike's point that you're probably not going to see otherwise because it's in a placebo-controlled large multi-center trial okay um now you're you mentioned the promotion to chief commercial officer of one of your BD people.

So what have you, what are you learning or have learned, you know, recently in terms of the marketplace that will help you with launching the, launching the drug?

Yeah, we'll likely be more elaborate as we get into Q3 earnings to talk about some of the specific research. But I think just listening into some of these conversations with KOLs and payers is clearly in non-obstructive HCM there's a huge opportunity and the profile of soda as presented is you know in layman's terms a no-brainer given you know the lack of other options given the relative low bar that's been set for us to be successful there what has been interesting is that there's still quite a lot of enthusiasm in obstructive HCM as well for various reasons. One, because despite alleviating obstruction, there's still a high degree of symptomology. Of course, dependent on how the drug functions in obstructive HCM, the safety and the tolerability of the drug still lends itself either as an add-on to a CMI in those particular cases or even in first-line use because of the ease of use of that particular drug. So, you know, it's been very encouraging so far and we'll present a lot more specific information on that at earnings. Okay.

And just one thing. So the data is 1Q27, but given you now know you're done with the enrollment, right? Do you have, is there, can't you be a little more granular with the timing in 1Q or are you still just going to keep it open to in 1Q or do you have a sense as to where in the quarter we might get the data?

That's a good question. Look, I think it's probably still a little bit early. You know, the closing a trial, it actually involves a lot of work. So you've got to get all the data in. You've got to make sure it's clean. Clearly, you do the analyses, et cetera. I think we'll be more confident to give a sort of more specific guidance at the start of next year.

Okay. All right. Fair enough. All right. Let's make sure we talk about some other things. We still have to talk about soda, but in a different area. So type one, where we, you know, I remember being at that adcom many, many years ago. But so things have progressed a long time, a long way since then. So tell us what we're waiting for. I know there's this very important natural history study, which is being conducted over in Denmark, I believe. So tell us what the significance of that is, this Steno1, why that's so important. And then, you know, just sort of thinking about what the FDA wants to see and the resubmission path.

Yeah, maybe let me start and Greg can dive into some of the details. So, you know, the FDA in the history of soda in T1D has really asked for a couple of things the whole time. And this time round, we feel very confident, in fact, having very good engagement with the FDA that we can deliver that. And those two things are we want a new prospective study, not a retrospective look at the data, cutting it in a different way. and we need to see from that study a certain amount of exposure on soda and to see a dka level that's acceptable and until now we've you know as a company we've never really known what those expectations are what we have now is we have that prospective study steno one that you're saying is running in denmark we have agreement from the fda that that study is sufficient and will gather sufficient exposure which they've given us the exact number and in fact we achieve that exposure at the end of august so the the data is done data is done now we just are in the throes of preparing it and getting ready to submit and we have a dka level that is less than what we saw in the in tandem program so it's all of those three aspects have been fulfilled We're in constant engagement with the FDA to prepare this resubmission. And they are the things that are different this time around. And so we are highly confident that we can finally bring this drug to patients with T1D.

So that sounds like some new information there, the fact that you've reached that exposure target at the end of August. Right. okay um what off the press just for you thank you so um uh so then in terms of the the dka rates are you is the point that you're those rates now are more in line with what would be expected in the t1d population or they're more because the steno one trial it's not just soda right it's a whole bunch of things if i'm not mistaken so how does it how are those dk rates comparing to you know all those other cohorts that are being studied? Because I think it's, wasn't it like 2,000 patients or something? It's a big...

Yeah, so you're exactly right, Igor, is it's a 2,000 patient study, and it's a treatment algorithm study. And what that means is patients were randomized by study site, either have a standard of care treatment, which was optimized lipids, optimize A1C, and then continue with the standard of care, or an enhanced treatment regimen, that the added treatment would be according to their baseline demographics. And the three options were either a soda group, a semaglutide group, or a fenarinone group. And as patients went through in that enhanced treatment group, the clinicians could add other medications onto that within that mix, depending upon their clinical judgment and the background patient characteristics. But there's a period of time when they're on one of those three. So you've got a thousand patients that are in this enhanced treatment group and a significant fraction of that thousand is on SOTA. And that speaks to Mike's point. It was just a matter of fully enrolling that thousand patient cohort, the percentage of those patients that are on SOTA, and then the accumulated exposure that met the threshold that we integrated with the FDA, which was roughly the total SOTA exposure in the in tandem registration program.

Okay.

The comparisons then we can make are we can make a DKA comparison of the SOTA group in Stena 1 compared to the standard of care group in Stena 1 and the rate achieved with SOTA with DKA in the SOTA group in Stena 1 compared to the historic rate achieved or realized with SOTA glifosin in in tandem. The primary metric that FDA set was comparing yourself to in tandem, because in their communication with us, which I think was in the end of review meeting that has been made public, is they accepted the efficacy. The only thing they wanted to get was a risk benefit right, and they said, if you get a prospective group of patients where that DKA rate is lower than that and in tandem, then the risk benefit will be favorable and the drug approvable. But just to be sure, we want to have both the rate compared to in tandem, but also the rate compared to the standard of care. And I think we've publicly disclosed this at earnings, that the rate that we've observed is similar to that of the standard of care in the steno trial.

So you would sort of achieve, if I'm not mistaken, and sort of a higher bar than even what the FDA was stipulating is a little bit, maybe.

I think it's just supportive. I think it's supportive from, we've certainly achieved the expectation that FDA has given us. And we've got a number of different angles that support the same conclusion.

Okay. And so now just very quickly, what are the gating factors now to get this? I mean, this just happened a few days ago, the exposure target. So when's the submission timeline? I mean, all that.

So there's been more than a nine-month process we've been working with Steno and the FDA on this. And we've had repeated meetings with the FDA. We've had ongoing, almost daily dialogue with Steno. We're at the point now of collecting all the data in the database that they use, all of the relevant metrics that FDA wanted to see. Steno is in the process of cleaning that data adequately for us to then take that database, put the final analyses together, and submit to the FDA. So that is the ongoing process, literally, as we speak, is that they are collecting the data, cleaning it to make sure that it is accurate, providing us all of that raw data in a database. And we've already covered all of the issues around patient privacy and all these other things, which again is a long process knowing the data privacy laws between the US and Europe.

And then we're in the process now of putting that in the final format for submission of what FDA wants. so you're at this point you're not quite ready to give a timeline on that we we've said q4 so it's going to be this year um and uh again you know as we march forward over the over the coming month we may be able to provide a bit more granularity at earnings okay um very good so then let's just switch over then to the the uh the dpnp program so there uh you know you had some very very strong data in the phase two and you determined a dose um you know you've been you've been talking about a partner um but just i guess just update us there because you know the emphasis

on that seems to be a little bit deprioritized perhaps no absolutely egal i think uh you know a number of things have transpired in the earlier part of this year uh that means you know we have we continue to progress that particular asset but HCM has risen to the top of our focus both from our resourcing as well as our sort of investor engagement etc and where the sort of nearer stage opportunity large opportunity is. Nevertheless one of the things that we we said when we were at that time engaging with potential partners as well as publicly is this mechanism AAK1 has a lot of promise well beyond dpnp and so to maximize the value of any partnership we wanted to ensure that they saw that value they could realize well beyond just neuropathic pain and we don't have the data ready for prime time yet but this year later this year we will present data, preclinical data, admittedly, but again with pilovapidin, the lead asset that we use in DPNP that shows really remarkable results in a number of new indications. And so we want to bundle all of that up as we look to progress pilovapidin generally and for DPNP specifically.

Okay.

So you want to give any teasers there in terms of what are your indications? nice try you'll hear it first mate along with everyone else okay fair enough um all right but i think it's fair to say these are broadly also you know aligned with where we are as a company in the cardiometabolic space um in areas of of high interest can i ask this i gather or maybe i'm interpreting that that some of the partners were saying like we'd be interested in x y and z and then then you did some of these preclinical experiments to get a little clue or or it wasn't like that uh no we look we and you know craig really instigated this uh early last year i'd say that we we have as you know through genome 5000 a pretty significant database of lots of different um studies preclinical studies across lots of indications for many many mechanisms and we went back and did a huge review of that and so we kick-started that work before we were really getting serious with the conversations with partners now those conversations with partners and some of this pre-clinical work were ongoing in parallel but certainly those conversations helped reinforce the idea that you know really let's make sure we capture the full value uh that this mechanism potentially brings well beyond you know what is a significant opportunity in dpnp but we want to make sure that um you know we're not missing any value by leaving potential uh high value indications on the table i think the other element and again i really thank you mike and scott for this is that by raising other source of capital we were able to do the capital raise, the deal with Novo, we took some of the pressure off of having to do a deal to finance, in part, finance the company.

So we thought that there's a much better value proposition of trying to get additional data and really validating AAK1 as a pathway. Just like we validated sodoglifosin as a pathway, we validated 9851 as a pathway, as we share with you, not only in obesity, but also in MASH and potentially in looking at lipids and cardiac plaque to maximize the value proposition. These are really interesting pathways where we're first, and in many cases only, and I think doing the additional work to have a disproportionate impact on value is what has been afforded with being able to find other ways to finance the company.

Okay, that's a perfectly sensible way to go. So then last one, so the Novo partnership, I remember at the R&D day in New York many years ago, you showcased the initial mouse data there, which looked pretty interesting. So ACSL5, you know, obesity is obviously super competitive space and dominated by, you know, So tell us why ACSL5 and what do you sort of expect to Novo to do with this drug?

Again, maybe I'll speak generally, and Craig, you can speak about ACSL5. But, you know, this was really one of the first things we did in my tenure as we looked to focus the company and focus our resources. We knew that we had in LX9851 a very powerful, as you mentioned, the results are pretty incredible. on weight loss either as monotherapy or combination on top of semaglutide but yet we had a number of other things that were sort of later stage where we wanted to focus our resources and this would be best served in one of the heavyweights hands and you know we had discussions with a number of companies and Novo really rose to the top because of a few things of course the financial aspect and This was a very lucrative deal for a preclinical asset, I think. But really, they were very enthusiastic about the mechanism. Craig can talk about that in a second. But then the other very compelling part of Novo taking this on is that, as you know, all of our medicines are once a day orally dosed. And we had the feeling slash knew that this would move to an oral-based therapy program, Like many cardiometabolic diseases where you use different mechanisms of action, pills in combination, and that's indeed where Novo thinks they're going. And they're putting this right at the top of their mission critical status and are moving at a speed of knots.

Yeah, the ACSL-5 mechanism is very important, early gate in triglyceride metabolism, both biosynthesis and degradation. And that's why it's expressed in the organs where triglycerides are most actively managed and packaged in the body, which is in the gut and in the liver. Now, triglycerides are really important in a number of major diseases. It's the major energy source for the body is triglyceride, and how the body moves triglyceride from the gut to the liver to the muscles and the other tissues is via these particles, lipoprotein particles that are primarily triglyceride. So having a mechanism that impacts triglyceride metabolism has massive benefit potentially in obesity, in MASH, which again is the mismanaging of lipids, primarily triglyceride in the liver, and then also in plaque because as you've seen with all these LDL and HDL and LP little a, these are all packaging addresses of these lipoproteins that go through the bloodstream. and if they're not packaged correctly, they get deposited in the bloodstream, which leads to development of plaque and cardiac arrest and in stroke. So I think that's why Novo has been so interested. While the primary focus is on obesity, there are these other potential benefits. And just to put a point on what Mike said, they only highlight a few of their early-stage assets.

Every time now, this is one of the assets that they highlight in their public statements, in their corporate deck so we're really pleased with having them as a partner okay um great uh one more just because we are asking everyone about ai very quickly if anyone wanted to throw in their two cents on on where you how are you using ai internally or yeah so uh you know at at the moment we're evaluating ai for what i think could be a pretty transformational program for the company uh and it's interesting we've never talked about this but you know it's not uh not necessarily uh material yet to the company but as you can imagine the genome 5000 database is a huge data set across both structured and non-structured data that was collected over the span of a decade or so uh and researching and extracting information from that database is very complex. It was set up 30 years ago. And we are now looking to how do we overlay an AI engine over the top of that, where we can make search queries automatic to enable not only work on our sort of current core programs, but if partners are interested in oncology, for example, we don't even really touch the oncology space of Genome 5000.

We have the potential to interrogate that database in a very automated fast and systemic way systematic way i should say and just to remind everyone that was something that was started i think in like 1995 or something like that and your memory is that that was the where you basically serially knocked out every gene in the mouse and tried to interrogate the phenotype yes is that that what it was 5,000.

Across, you know, 60 or 70 different sort of therapeutic trials, preclinical trials across oncology, metabolism, cardiovascular disease, immunology, etc. And, you know, all of the targets that Craig and I have discussed today came out of that program. All of our medicines are internal. We haven't in-licensed anything. It shows you the power of that genome 5,000, But really, given the sort of dating on it, there's untapped potential that we're not necessarily taking advantage of completely. And we think AI will provide us the tool to do that.

Very interesting. We'll have to wrap it up there. But thank you all very, very much. We appreciate you.

Thanks, Igar.