Twilio Inc. (TWLO) Earnings Call Transcript & Summary

All right. Welcome, everybody. I'm going to say the Morgan Stanley disclosures. For Morgan Stanley research disclosures, please see morganstanley.com/researchdisclosures. I'm Meta Marshall. I cover communications software here at Morgan Stanley. We're delighted to have Twilio here. Delighted to have Khozema Shipchandler here, who is CEO. You've had all the roles within Twilio, but your current one is CEO. So...

Thanks for having me.

Maybe start with Twilio has changed pretty meaningfully since you took were as CEO, and since you've kind of been -- come to the company in general. You've done some headcount reductions, business reorganizations, change go-to-market motions, most of which you covered at your latest Analyst Day. Can you just outline what investors should take away from the Analyst Day and how the Twilio vision has changed alongside of that, if at all?

Yes. I think first of all, what I would say is, is that we were just talking as we're coming on stage that last year at the same time, during the conference, we talked a lot about the financial discipline with which we wanted to run the company, the operating rigor and the innovation focus. And I think in the year that's transpired, you've seen like a number of proof points that sort of validate that. And as you look at our Investor Day, I mean the past is kind of the past. But in that period of time, we took a company basically that was fundamentally losing money that was using cash whose growth had stalled. And every one of those metrics in a kind of different trajectory, but every one of those metrics is starting to turn around. And so I think the evidence of that is starting to show up. And then from our standpoint, like we're incredibly excited about the vision that we painted for customers going forward. And I think that's a great setup for investors as well, right, $3 billion of cash flow that we've committed to over the next several years. We definitely do see positive signs in terms of the growth trajectory of the company. We've given guidance that we feel good about. But fundamentally, we're running it to something higher than that. And we think that whatever that higher number is, like all of that's going to fall to the bottom, and so investors will benefit from that, too. And then coupled with all of that is another authorization on share buyback. And so I think just the way that we're running the company, Meta, it's a lot different. And I think from our standpoint now, the discipline and the rigor never stop, okay? That's -- those are forever elements. But going forward, it's about how do we start to focus more on innovation as a means to kind of set up the next 3 to 5 years.

Perfect. So at your Analyst Day, you laid out a TAM of nearly $120 billion in your core markets, growing to nearly 160 layering in AI. Can you just outline for investors the various growth drivers within that TAM? It's a big TAM. And just why Twilio is now better positioned competitively to capture that opportunity?

Yes. So I think the shift in the TAM that we see happening, I mean it's a big TAM as it was. And if nothing else changed, let's say, I think our ability to continue growing within a growing TAM between now and 2028, let's just say, just in CPaaS. I think that would be a pretty decent story. And our ability as a company to have taken share relative to our other competitors in that marketplace. I think that was going to be there anyway. And so that continues, right? So that's one. I think the second thing is, is that the intersection of the categories that we're seeing, right, with CDP, with CCaaS, with our core CPaaS business, the customer experience as a service category that's opening up. And it's not just what an analyst writes about or what we're trying to say, like, this really manifests in the actual customer conversations that we're having. And as we look towards that, like we see contextual data is the thing that's going to ultimately activate that growth trajectory. But more tactically, I mean, there's a lot that we're doing on self-serve that's been really exciting over the last 12 months. There's a lot that we think we can do more of with ISVs, which continues to be a great part of our story. And then I'd broaden that out to kind of partners of all kinds. I'd say partners is an area, frankly, that we've not done as great of a job in. And I think there's obviously volume leverage distribution there. And then international. I mean we're underexposed as a company internationally. And so I think there's a combination of just natural market growth levers that we are going to activate ourselves that I think are totally a self-help story and then this market convergence just opens up an even broader TAM.

Okay. So kind of on that market convergence, one of the largest value propositions for Twilio has been the ability to combine communications and contextual data versus with segment. The cross-sell motions over time have kind of underwhelmed maybe. Can you help outline what has been limiting those opportunities and how the changes you've made can kind of get us to achieving more of that [indiscernible]

So I mean, I think your question is right. I mean, I think we've done historically 2 things relatively poorly, right? I think one is that the way in which we organize our go-to-market team. We went through a number of different changes there was fundamentally flawed. And so refocusing and having dedicated sales reps that can just take CDP to market because you got to seed it to be able to kind of unlock additional opportunities going forward. But I think, honestly, the bigger limiter was we never actually launched a native product with which you could activate data inside of a communications API. So one of the big things, and we talked a lot about this last year as a result of our segment operational review is at some point, like you have to listen to customers in terms of what they're telling you. I think having native data capabilities through a communications API, that was paramount. We saw a bunch of success with that. We're now starting to combine that with AI workloads, too, which I'm sure we'll talk about. I think that's been a real unlock in terms of customer conversations because Twilio has always been like the easy button company, right? And a customer wants to be able to come to the platform. They want to be able to access an API, they want to be able to consume data, and they want that to be super simple. So I think all of that is sort of in place now. And then with respect to go-to-market, what we have done is we've kind of functionalized our go-to-market team. However, we've kept the segment sales reps a little bit isolated. We put much heavier incentives in terms of cross-selling. And I think the combination of having products, having kind of better sales execution, and we've done a lot over the last year that I think is validated that, too. You see some of that, obviously, on our balance sheet. That all supports the thesis that there's a thing here, and we're executing against it, and we see more opportunity.

Got it. So you mentioned ISV and self-serve have been go-to-market motions that have started to kind of see more traction in. We saw this evidenced in kind of the double-digit growth in Q4. Can you just speak to some of the initiatives that helped drive success here? And you mentioned kind of partnerships being another area where you guys still have some work to do. Like just what does that addition of additional kind of partners for that flywheel?

Sure. So to start on the self-serve side, I mean, obviously, this is like the heritage of the company, right? And so in some ways, it was sort of like a back to the future moment for us. In that, we'd always had like a really strong brand. We'd always had a strong self-serve capability. I think inside the company, the way that we characterized it was we wanted customers to be able to get to a magic moment in minutes. And over the last couple of years for reasons that are, I think, necessary to protect the ecosystem, working with a number of our carrier partners, we've introduced a lot of friction into the process of just signing up an account, like putting your use cases out there in the world. And so honestly, a lot of the work was technology work behind the scenes that's like kind of hard to see for the -- maybe the average person, but very apparent to developers, right? So when a developer attaches themselves now, like a lot of that friction is gone, right? And their ability to seamlessly go through the self-serve sign-up process is made a lot easier. Their ability to access multiple products is a lot easier. Their ability to understand exactly what's happening when they communicate with customers, it's like they're in dashboards through analytics in the self-serve platform. So just a lot of things behind the scenes that we did to better activate people that were coming in through that funnel. On the entirely flip side of the equation with respect to ISVs and partners, I think one thing people don't fully understand about ISVs is like there's some really, really big ISVs that we're obviously super proud to call partners, some of whom will be presenting at your conference over the next few days. But actually, the tail on ISVs goes into the thousands, right? And so not only are the growth characteristics attractive actually, the gross margin characteristics are actually better than the corporate average in spite of what you might think relative to some of the big ones that are able to, obviously, negotiate price as a result of volume. But I think the one thing that we've never unlocked with ISVs and the same concept extends to partners is with ISV specifically, we don't co-sell with a lot of them, right? Like we're not really competing, like the only real product that we are ever competing in directly, we pulled away from Engage Premier, which you're familiar with. And so now there should be no friction in our ability to work with one another to go pursue customers together. And look, I mean, there's always going to be some level of coopetition, I guess. But ultimately, our infrastructure is required. Ultimately, we're not competing with them on the Martec side. And so to the extent that the ISV is Martec oriented, and I think that presents an opportunity. With partners more broadly, what I would say is as you get into these implementations where it's communications, data and AI, if you're Morgan Stanley, like you probably have the resources the technical wherewithal to be able to string things together on your own, giving your own in-house expertise, your own developers, your IT teams are very sophisticated, but that may not be true for the average retailer. And so enter SIs who now have an ability to make a buck off of stringing these different things together. That's not business we want to take on. We want to push the building blocks out, and I think that's a win for everybody.

Okay. In terms of balancing kind of developer-led motions or kind of your self-serve heritage, where do you see kind of enterprise sales entering into that? And just what kind of efforts are you making to improve that enterprise motion?

I think it's a couple of things. So one is the enterprise is still -- like they dominate the revenue line, right, in many respects. And again, like with 325,000-plus customers, clearly, a huge number of those are large enterprises, they tend to be the bigger spenders as well. I think the message that's really resonating with them, and it goes to the kind of largest deal that we announced in segment's history in Q4 to a number of other really exciting customer conversations. These are companies who are very excited about the vision of communications, data and AI. Like they want to create like outsized ROI, whether that's both on the bottom line with respect to reducing their cost basis in terms of like AI agents and stuff like that, but also the ability to increase their revenue line. And so I think the misnomer, perhaps inside of Twilio before was it takes this like super consultative, like specialized sales rep. And I think that's not true actually for us at least. And I think instead, we can lead with developers, a lot of these enterprises come in actually through developers, and we can kind of build those relationships over time. Ultimately, like everyone of our customers, like the only thing they really care about is ROI. And if we can prove ROI to enterprise customers in the same way that we can to everybody else, like that's a really compelling story. Getting in the door is not really the hard part, like ensuring that we have compelling ROI in front of them, like that's the thing that we've really got to train the reps up for. And then I think, again, as I said, I think partners can play a really crucial role.

Okay. So you alluded to kind of the AI opportunity. At your Analyst Day, you kind of outlined this $260 million of revenue that you've had over the last 12 months from these customers. Just how does AI unlock contextual data and your broader value proposition? And where are you kind of seeing that success to date?

Yes. So there's kind of two things in there, I guess. One is with respect to AI. I mean, I think the opportunities in AI are obviously amazing. The technology is incredible, and we're really excited. I think with respect to -- we quoted a few numbers, right, $260 million in revenue, 9,000 companies, 90% of the Forbes 50. Like at some level, no matter where you are in the Twilio life cycle, like all of it is attractive to us, right? It could be as simple as an AI start-up who's activating against our infrastructure, which could be voice capabilities and they're getting going that way. It could be even simpler. I mean it could be that they need some authentication capability, and they're coming to Twilio. And so we're sending out a bunch of 2FA or SMS as notifications on their behalf. Obviously, the kind of higher order stuff is companies that are coming in and using a variety of capabilities, right, so that communications data, the AI or the contextual data and then AI, all of the mix. I think ultimately, like that's what we believe is sort of the destiny of the company. I think it's validated in the customer conversations that we've been having. I think the reason we're so bullish on it, Meta, is because in all of the customer conversations that we have, yes, companies are very excited about AI writ large. I think there's a continuum, right? So you're going to have let's say, nonregulated industries like food delivery, restaurants, retail, e-commerce, like the stakes are just lower than in health care or financial services, right, or insurance. So I think that maybe drives the adoption cycle in some ways. Although, I was at a health care conference last week, and I was a little surprised by how quickly folks are ready. But anyway, I think the key point is, is that companies will never give up their contextual data. Like it would be crazy for them to give up their contextual data to help train up large language models as amazing as they are to be able to disintermediate them from their consumers 1 day. Like that's just not going to happen, right? And I think you see that also -- it's not just Twilio saying that, I think you see that through the lens of the big cloud providers. You see that through the lens of the big data warehouse providers, like the CISO has a lot of power right now, right? Like they do not want data to leave their environment. If anything, they want a platform like ours to interoperate with their environment to then -- but to keep the data there, right, to be able to drive value outside. And while an LLM may integrate, the data stays there. And whatever the training is that's done stays in the customers' operating environment doesn't leave. And for obvious reasons, like you wouldn't want -- if you're a retailer A versus retailer B, you probably don't want a model being trained up so that you can sell more stuff over to retailer B, right? So I think it's sort of intuitively obvious, but now we're hearing it like time and time and time again inside of our customer environments.

Got it. I mean how do you see yourselves as able to really establish ourselves as kind of that vendor agnostic platform for AI names to reach customers versus trying to kind of offer something on your own?

Yes. I mean, I think, again, it speaks to using the strengths that we have, like having 325,000-plus customers, that's a pretty good starting point, right, to grow off of, to cross-sell into, to upsell into, again, I think there's sort of like a continuum here in terms of our revenue stack, right? A lot of it in the early days, like if we just sold more Verify, which has very high ROI, like very protective to our customers, like that's a relatively straightforward sale. If we did more Voice Intelligence, again, very high ROI uses, let's call it, AI light capabilities. Like I think there's a progression, right? Not everybody is going to go necessarily for a big bang, and we don't need them to necessarily. So that's kind of through the lens of like how we sell through the lens of like how customers are buying, I mean, I'll just -- I'll give you an example that's relatively recent from the health care conference, like we ran a demo that we put together in 2 days, okay, in a hackathon, this is all public. And basically, what we did is that we developed an AI agent, again, using contextual data about health care consumers that's basically able to solve billing inquiries as they come in, okay? Now for all industries, about 80% of all call volume is driven by billing, okay? For the health care -- I was actually -- myself astounded by the statistic. For the health care industry, it's 97%. Okay? So now you've got an AI agent who's able to not just drive the interaction, right? So that's kind of the flip from like IVR to an agent, right, is, okay, it's a one-for-one replace, but it's driving the interaction. But in addition to that, no queue, right, that's pretty attractive, if you're a customer. But most importantly, able to solve the problem in real time, right? Using data, the AI agent knows about that specific consumer. Like that's hard, right? But -- and it's not possible without one's ability to actually use contextual data as the means to drive that. And I'll just tell you a funny story as I was leaving that particular conference, like one of the health care CEOs approached me and said, hey, listen, my number one issue in my business is people not showing up for appointments on time, okay? And so he said, do you guys think you could help us send a text message that would like remind them to show up for their appointments? And I was like, we have been doing that for 16 years, okay? So -- but it's just -- I'm not like being flipping it, like it gives you the range of opportunities to be able to unlock ROI in these various industries it doesn't have to be like all AI, all the time, all-in. Again, if you look at it through the lens of ROI, there's like growth from that relatively simple example of notifications, which is real ROI for them, right? If customers are showing up on time all the way through to these more sort of agentic offerings. I think it's really cool. I mean it's super exciting to be in technology right now, frankly.

Yes. I mean on that those are two great cases where you have a solution today for a customer that maybe you didn't even know that you were the person who could provide that solution. And so as organizations turn to system integrators more particularly when it comes to AI, and they're a little bit more like I don't know where to start, what can you do to kind of improve relationships with system integrators to make sure that people know that you have data solutions?

Yes, great question. So I mean we have a good brand. That's a good starting plan, right? So I think if you're a systems integrator, like what you know at some level -- some core level is, is that you need communications infrastructure, right? Typically, I think we find that, that value would accrue to us just because we're a well-known brand. That said, I think it's on us, though, to sort of like really invigorate -- reinvigorate our partners program, right? So we have hired a new head of our partners' programs. He's kind of a seasoned veteran, having done partnerships in a variety of industries, predominantly in technology over the course of his career. I've been very excited about what I've seen so far in terms of just like these early relationships. But again, I think we've got to be careful about like not overpromising from the word go. Like, again, I think it's this continuum of things. We've got to make sure there's a set of capabilities that are building blocks from our side that they can then string together so that they can see their way to making money off of Twilio by doing the integration. And I think historically, we've done a great job of that, and there's a lot of at least effort going towards that today.

Okay. So maybe diving back into kind of the reacceleration that you've seen -- you have seen reacceleration past 2 quarters and particularly a pickup in deals kind of greater than $500,000. Just what are the keys to kind of sustaining this recovery that you've seen? And kind of what drivers do you think ultimately kind of got you that reacceleration?

Yes. I mean I think a lot of it is honestly unlocked through innovation. I mean I think just to unpack it sort of fully, the reason we've been excited about what we initially called stability and that what we've seen since it's become reacceleration is that it's pretty spread out. Like there's no like idiosyncratic factor like social or crypto or something like that, that's really driving it. Instead, we're seeing strength across industries, I think, which is pretty encouraging for our overall business. So that's kind of the starting point. I think the second thing is, as we showed during our Investor Day, there's a number of these new features that are pretty accretive to our top line, like a Verify, like a Voice Intelligence, but also like very ROI enhancing for our customers, right? So it's like truly like a win-win, right? Our customers -- they actually, in many cases, get to ROI in like 30 days. And for us, like these are upsell capabilities that add price and obviously accelerate revenue volumes. As I mentioned, we put a lot of investments into self-serve. We've kind of equally put maybe not technology investments, but although some with ISVs actually, but also kind of human investments and partners, as I mentioned a moment ago, I think that helps, right, kind of provide a little bit more wind in our sales. But I think ultimately, like long term, this is going to be an innovation story, right? And I think our ability to make it easy button simple for customers to avail themselves of data while they're using communications workloads, like that's going to be the story sort of over the medium term. But in the short term, it's all self-help. Like it is a go-to-market execution story. It's predominantly cross-sell. You saw the stat right at our Investor Day that people that -- customers that use more than one product drive 90% of our revenue, yet the majority of our customers don't, right? Like I'd say that's a huge opportunity for us. And what's great about it is, again, we're very aligned to customers. Like when you're using 2 products, you're not doing it for fun, you're doing it because like you get ROI from that. And so I think putting it on us to drive an ROI story, that's a game we're happy to play all day long.

Got it. And some of the questions that were old and new again maybe on the RCS messaging, which seems like a question we've been asking since 2019. How has that messaging platform evolved since its beta launch? And just how do you think about kind of the threat, the opportunity that kind of comes with RCS?

Yes, I don't think it's a threat. I just -- I'm sort of like neutral to kind of mildly optimistic about it. Like I mean, I'd love for it to take off, frankly, okay? I think we have a great partnership with Google. And so I think there's a lot to be excited about there. I think that at least today, I see a little bit more limited utility, right? And so I think some of the reasons for that are like for notifications or authentication, like you probably wouldn't use RCS, right? I mean it would be overweight for those kinds of use cases. I think an example I've used some time is like you travel all the time, right? I travel all the time, like -- but a lot of that is actually driven in the app, right? And the airlines have worked hard to get you in the app. They're not about to have you leave the app. Like, I mean, they have -- you've given up valuable real estate on your device for them to kind of dictate the consumer relationship. That's really, really attractive to them. They're not going to flip it over to another place, certainly not to give up data, I think, which I think is a threat. I think what you will put up with from the airlines -- and they do -- most of them do this today, you'll put up with SMS messages because they need you to know if something like went wrong or is going right, hopefully, most of the time. Similarly, e-mails, right? Like they'll send you a lot of e-mails to kind of confirm choices to confirm seating, meals, stuff like that. So I think it just has kind of more limited utility in that way. And then I think structurally, there are some interoperability issues still, right, with respect to Apple. So I'm cautiously optimistic, but today, at least I see it as kind of a neutral.

All right. Like they're only going to tell me bad news and then text and then maybe the in-app tell me good news upgrade and...

Well, I don't know. I mean it's -- I mean, just not to -- I know you're making a joke, but I think the other side of it is like think about it, okay? If you're in the app, like I think what's really interesting for us and should be for them, too, is like they have a lot of information about you inside the app, right? And we already do this for some airlines. Like let's say, God forbid, your flight is canceled, right? There's a -- you're a frequent traveler. It would be so simple for them to like just push you a coupon to say, hey, look, thanks for your -- we appreciate your loyalty. We apologize for the trouble that we've caused you. Here's a coupon to go visit our lounge like have a few drinks on us for the hardship that we're putting you through. Like that wouldn't be so complicated, right? Like -- and again, using contextual data that they have for you, they can even create like a rank ordering of who those customers are, but it requires actual data. Now with them, they have the ability to roll that inside the app. And for Twilio, like this is awesome because it combines like what we were just talking about: Communications, data and then you use a little bit of AI, not a lot, a little bit of AI to like create a better experience. And you can take what's going to be otherwise a terrible experience for the customer. No one likes a flight cancellation and just make it better, incrementally, right, incrementally better. And these things work, actually.

Okay. That's a great example. I look forward to incorporating that. So...

I think you'll see it soon actually.

Okay. Perfect. So you've seen acceleration in the business over the last 2 quarters, as we just mentioned. So maybe we're a little bit more cautious in terms of your 2025 outlook than investors were expecting, particularly after Q4 earnings. Just how do you look at kind of the guidance philosophy and just what factors could cause the strength you saw exiting the year not continue into the next year?

Yes. I mean I understand the question, okay? But we also put out our earnings about 3 weeks after we just put up the framework, right? So I am a little surprised that anyone was surprised that we would -- that we would somehow like alter all of our guidance all over again. So look, I think we're a couple of months in. We feel pretty good about the kind of tailwind that we saw in Q4. We certainly feel good about the setup for the year. I don't think our philosophy around guidance has really changed all that much. I mean, you know the company better than most, like the usage dynamics that kind of dictate a little bit of the way that we have to guide. I'd say it's about the most dynamic macro environment that I've seen, certainly in a couple of years. And it's not more complicated than that, right? We are running the company towards double-digit growth like we feel good about the things that we're doing to be able to kind of get there over time and whether it's go-to-market, whether it's product investments and like we want this to be an execution story, not just one of the guidance.

Okay. So another highlight of the company over the past couple of years has been profitability. You guys had your first quarter of GAAP profit profitability last quarter. You've been operating or orienting the company towards double-digit growth. But can you just highlight for investors kind of what kind of margin upside do you think is in the model with double-digit growth?

Well, look, I mean, I think we...

Or potential.

Yes. I mean, look, I think we've said 500 to 600 basis points, right, of margin accretion over the next few years. We certainly feel good about that, right? So I'm not going to change the guidance on stage and create news. But I think what I would say is we feel very good about our head count, okay? We don't see a need to like dramatically increase the number of resources. I think on the contrary, we've seen really, really good productivity gains inside the company as it relates to support and SDRs in particular. And so right now, we're like quite focused on taking what we've learned there and driving it through the rest of the organization. Anywhere there's kind of repetitive work, I think it's an opportunity or there's some internal facing agent, so to speak. So I think if you just kind of run the math, like if we get there, and we feel pretty good about our head count, as I said, you should see upside if we achieve those results. And as we've been kind of consistent about, that will kind of drop to the bottom line. I think the reason we're bullish about the 500 to 600 is we are just starting with some of these AI automation investments. We're just starting with respect to like what we characterize as workforce planning. And then I think there's another side to the story, which you alluded to with respect to GAAP, which is like we've also done a lot of change on the equity line, right, just in terms of really managing burn, kind of putting aside a stock buyback like really managing the burn equation, like, obviously, like that creates kind of change for employees as well. And I think employees have been remarkably engaged. Obviously, stock performance helps. But I think they want to be a part of the story over the next 3 to 5 years. And so they've really hung in there and our attritions remain relatively low.

I mean M&A would be kind of a bigger part of [ Twilio. ] You guys haven't done as much. You just talked about the $3 billion you're going to generate over the next few years. You've talked about kind of share repurchases. Just how are you thinking about M&A as kind of part of your strategy?

Yes. I mean it's in there. I think to be honest, I mean, like we have sort of a mixed track record on M&A. And so we're very circumspect about where we put that next dollar of M&A investment. I think when we've been adjacent, we've been a lot more successful. I think segment, we've sort of unpacked the lessons learned from that particular experience, and we know what we would do differently if we do all over again. I think we're running sort of pre-mortems, if you will, to understand what the traps could be. But look, I think, for us, it's predominantly going to be like tech tuck-ins and stuff like that. Like I wouldn't take from that, that there's like anything imminent plan. There's no signal I'm trying to send there. It's kind of early in terms of M&A for us, but it's something that we're evaluating, and it's part of the overall capital allocation of the company.

Okay. And then just kind of last question, something we're asking everybody at this conference. Just how are you using AI internally? And have you had any demonstrated success from it?

Yes. I mean, I think the 2 biggest areas are in customer support and with respect to our SDRs. So in customer support, today, 75% of incoming tickets are deflected. That is a massive lift in productivity. So you can imagine, as a result of double-digit growth, you also see increases in ticket volumes. And in spite of those ticket volumes, we have not had to add any headcount. And in fact, quite the opposite, we're using AI as the mechanism to deflect those tickets. And most of those questions, I mean, a lot of them turned out to be billing, as I said a moment ago, those just get answered by a virtual agent, whether it's over the phone or via text message. To me -- so a lot of people are doing that. You're probably hearing that from a lot of companies. The more interesting one, from my perspective, frankly, is we took the exact same technology and said, this is all homegrown tech, by the way, too. So we took the exact same technology and said, all right, well, what if we turned it outwards so that as we get leads, we use those as a way to kind of vet leads, right? So the job of an SDR can be pretty laborious like having to sift through all this volume. And so in the case of inbound leads now 80% get deflected or I'd say, more appropriate words, filtered, right? So now the leads that and actual SDR is handling. So there's a productivity benefit. But the more important benefit I would point to is, is that it's actually a qualified lead, right, on the way in. So I think those are two really profound things that we've seen inside the company. We've done a lot with AI and ML, which I've spoken of before in terms of just accelerating like our onboarding process or compliance process, like that's taken work that used to take 3 months. It's about 3 seconds now. So that's pretty powerful. So I think the idea is like taking some of these things now and spreading them to other departments, but yes, we're seeing pretty good results. And we're kind of contemplating do we productize this stuff over time.

All right. Perfect. Well, Khozema, very exciting to have you here today, and congrats on all you guys have accomplished over the past few years.

Thank you.