Q2 Holdings, Inc. ($QTWO)

Hello, and thank you all for joining us. I'm Josh Yankovich, Vice President of Investor Relations and FP&A at Q2. A brief note before we begin today's call. Some of the comments today may include forward-looking statements that are subject to risks, uncertainties and assumptions, which could change. Should any of these risks materialize or should our assumptions prove to be incorrect, actual company results or outcomes could differ materially from these forward-looking statements. A description of risks, uncertainties and assumptions and other factors that could affect our financial results or outcomes is included in our SEC filings. Except as required by law, we do not undertake any responsibility to update these forward-looking statements. All right. Today, I'm excited to host Adam Blue, our Chief Technology Officer; and Hima Mukkamala, our Chief Operating Officer. We've organized today's session as an AI-focused deep dive covering Q2's strategic position, how Adam and Hima think about AI as an opportunity and the technical and architectural advantages that differentiate us in financial services. This is a conversation we wanted to have for investors heading into earnings dedicated entirely to questions on AI.

So with that, let's go ahead and get started. So, Adam, we'll start with you. Every software company is being asked about what AI means for their business right now. I'd like to start with the opportunity side from a strategic perspective, what is the opportunity set that you see available to us at Q2?

Yes. I think the opportunity is fantastic. Throughout Q2's 20-year history, we have been a partner to banks and credit unions, when technology evolved and shifted to help them take advantage of that technology to be competitive within their base, in their market and with the largest money center banks. So when you have a real technology shock like the availability of generative AI and all of the attendant changes that have come as a result over the last 2, 3, 4 years of everyone getting their heads around it, that actually increases our value to the financial institutions that we serve because we have built a platform and deploy it with them that deeply understands what they do every day to operate the digital channel and their bank generally. And that platform represents a substantial site of embedded value, deep understanding and context that we can now leverage using AI to go faster and do more and accelerate and amplify the things we've been literally talking about for years. And so as an opportunity, I think it is 100% aligned with our existing business model. We're not really a SaaS provider that delivers a SaaS application that people sign up for and use and then can easily move to another application. We have some characteristics of our business that are very SaaS-led. But at the end of the day, we are operating a complex multi-tiered platform with a wide variety of crucial integration points in an Internet-facing completely open Wild West kind of environment, and we are helping banks and credit unions assemble that set of technology units in the face of market challenges to maximize how they go to market and how they compete. And so anything that arises from a technology perspective that raises the bar for everybody is really, really valuable to us.

I think it makes a lot of sense. Hima, anything you would add to that?

Yes. It is an exciting time what the technology is enabling us to do not only for us, but also for the customers. Like Adam described, context is the pillar, it's the advantage and AI is the amplifier. So that's the strategy that we use, that we continue to build on. Like, an example of it is we have announced a new product called Q2 Code. It lets bank developers build customization on top of the platform that is driven by using AI agents and AI assistant. So to do that, they have the context of their business, they have the context of how their rest of the application looks like, but then constrained within our security controls, our compliance, all of that infrastructure baked in to create this runtime that is secure. So they're not starting from 0 with AI. This is a continuation of the journey they've been on. We are seeing this internally how our velocity is getting accelerated. When a customer comes to us and says we need to deploy some capability in 90 days, we can now say, yes, we can deliver that. We have Q2 Code. We have the infrastructure that's in place that's been there for a long time. So the organizations that rely on us, our customers, they can move fast. But more essentially, they can sit in compliance with the rest of the environment as they go through this shift cycle. So that's what we're building towards.

Great. I think a little bit later, we'll come back to some of the cool things that we're doing internally and dig into that. But Hima, maybe I'll start with you on another kind of question we get from investors a lot, which is maybe it's a good opportunity for us to talk through another theme, which is understanding where Q2 sits in the architecture of banking and why our position is advantageous. So maybe Hima, we'll start with you about the integration, orchestration layer and why that is important with the application of AI.

Yes. I think I want to continue on what Adam said, right, which is Q2's digital banking is not a thin layer that is just visualizing a bunch of core data, right? As he said, we sit in all the customer conversation, customer workflows that are happening within a digital banking. And we do this with a multi-tiered architecture, and that is where we sit, right? As Adam said, there's a lot of complexity in the cores, how they operate, what are the nuances to it. So it's a rich integration layer that's needed to make that happen. But we need to orchestrate all the actions that the financial institutions or the end users take. And that orchestration is pretty complex in terms of wires, ACH and money movement as a good example of what it takes. And then finally, visualizing all of those. So the point I'm trying to make is we are not a thin veneer that is just taking a bunch of data and just showing that data without not having a context to it. So an example of this is the bank will come to us and say, "Oh, we're thinking of changing our core system. And you would think, oh, why are they coming to us? Because they are not just thinking of the core system, but they're thinking of the whole infrastructure when they're thinking about the core. And what we get out of that is what we hear and what the customers are saying is, before we do anything with that changing the core, thinking about what is needed to change the core, we need to make sure this rich integration, orchestration context system is ready for it, right? And so that they're putting a lot of emphasis on this architecture, the system at the top because that's where the customers live. So this digital layer has become like the key product. They're trying to protect first before they think of core migration. This is not a layer they're willing to change. They'll come to us and say, we want to change core. So how would you accommodate that versus coming at it digital changing versus core changing. So that's something which we hear a lot. And this is a continuation or a direct result of how much context and operational integration that Q2 is into this financial institutions operating model. So we are not just integrated into the core, we are a fundamental system that helps the financial institutions operate. We are embedded into how they think about risk, compliance. And yes, Adam, anything you would add?

Yes. I think your notion about system of context is really powerful here, Hima. The other thing I would point out is the digital channel is where the orchestration occurs. It's also effectively the system of orchestration. So when I log into digital banking, I'm going to see data from the core, data from the bill pay provider, data from Zelle, data from SavvyMoney to tell me what my credit score is. And then the digital banking layer also knows when I log in, what I do when I log in, what my preferences are, which accounts I like to see in what order, what my nicknames on those accounts are. None of these things are downstream in these other products. And so the experiential data and the expressed preference data for the end user, whether it's retail, commercial, treasury, all of that is up at the digital channel. And so what digital banking looks like is comprised of this integration of data from -- in a typical implementation, 25 to 35 distinct sources. And then the orchestration of, okay, when I do a payment, say, I got to make sure that there's enough money in the source account. If it's a prefunded payment, I execute the transfer from this Q2 platform, make sure the funds are there when the payment clears from the Fed. I'm checking entitlements. I'm making sure that there are security check. All of these things interoperate in a way that is really, really valuable and very, very complex. So there's an integration layer that provides that system under which we can bring that data together. There's a context layer that maps all of it in a reasonable way. And then there's an orchestration layer that really does what the end user wants without them having to know what's happening underneath. All of that plumbing is kind of invisible, which -- because we're very good at it. But the invisibility of that plumbing is where a lot of the value is. So our [indiscernible] architecture, I think, is more vertical inside the bank than a lot of people realize. And we're performing a lot more orchestration, a lot more context creation than is apparent probably from what you see in digital banking.

Great insight from both of you on that. And maybe we'll kind of segue that topic into another one that we have been hearing from investors around kind of a buy versus build dynamic. So a lot of questions from software investors is AI-assisted coding is making pushing out code easier than ever. How do we think about, like, our banks and our customers building things that we have built for them but doing it themselves internally. Adam, I'll start with you. Is that something that you're seeing? Or how do you think that will continue to play out over the future?

Yes. I think it's really interesting. Financial institutions like to build things, but they don't want to do the boring stuff. You know what I mean? Like if you had $6 million burning a hole in your pocket, you could probably go buy a really good condition used citation. But are you going to fly it? You're going to maintain the engines yourself? You do the telemetry? You can do your own air traffic control? You're going to put gas in it? And so owning the code that delivers that channel and taking care of it, maintaining it, extending it, adding capabilities, that is not fun times for anybody. And so when I think about our customers and buy versus build, what I tell them all the time is, let us do all the boring stuff. Let us do the hard stuff. Let us do all the Cloudflare integrations to make sure that it's secure. Let us deal with your regulators and correct coding practices and all those pieces. Let us deal with SDLC and engineers and product managers and all that stuff. You take Q2 Code in the SDK, you just build the fun parts. And the SDK is so powerful because it sits on top of an abstraction of their core and the data from bill pay and the data from payments and the data from integrations and security data. So every line of code you write or better yet have the AI write for you, it represents 10 or 20 or 30 lines of code in the underlying API infrastructure. So I guess a financial institution might wake up one day and say, let's just build our own digital banking channel. What I can tell you is, unless they're committed to putting together the level of talent you have in a Tier 1 digital banking organization, even if they do build it, they will be so far behind what the current state-of-the-art is by the time they finish the initial build, they will literally never catch up because we're going to start building features and extending the platform and extending the infrastructure at the accelerated velocity that AI enables. And so we are already hitting the ground running. And frankly, when I talk to financial institutions, they're coming off of other platforms from other participants in the market where you have everything you wanted, you just had to build everything you needed, and they want the exact opposite. They don't want to have to build anything they need. They want to build only the small pieces that they want on top of an extensible platform. So I think the AI revolution, if you will, and Q2 Code let us lean more into, hey, you guys build the fun stuff that really addresses your use case in your market and let us build the other 95% of the application because that, I think, is where they want to land.

Hima, anything you'd add before we go to the next question?

Yes. I mean, Adam referred to this, right? Vibe coding, Agentic assistant coding, whatever you want to call it, that is where the market is headed, software as a space, but Vibe operating is not something people really think about or know what it takes to go build 24/7, something happens. Someone wakes up middle of the night, 2 a.m. That's what the structure and the operating model is what Q2 brings. The reason why this question of buy versus build comes because are we able to move fast enough to meet their needs from their customers? And that's where AI, that's where Q2 Code, that's where our SDK Innovation Studio, either us directly or through our partners are enabling them to move faster and do the real business value outcomes, fun stuff, as Adam said it, and leave the other things, we are moving fast in terms of how we're releasing products in the market, how we enable customers to build those customizations or new capabilities that reflects their business. So that is where the whole notion of core versus context. We want the banks to focus on their core, which is running a bank and build the fun stuff on top of the platform, while we go leverage AI, expose AI, I think that's the key, right? Not only us leveraging it, but exposing it to them through the tools which we have so that they can build fast themselves on top of the platform.

That's great. And maybe that kind of dovetails into the next question I have with Hima, I'll start with you. When we talk to investors about the AI work underway, some of which we already shared publicly on some of our most recent earnings calls, what are some of the real tangible things that you're seeing out there today? And what are some of the advantages that you might see in the future going forward?

One of the most sort of big decisions we had to make was we don't have a separate AI team. To make AI successful for us as an organization, we did some early POCs. There's so much changing, but we made it a fundamental part of how we do our work today, right? And that has been a critical part of changing how an engineer thinks about it, a marketeer thinks about it or an implementation person thinks about it. And so that is forcing all of us to think and including me, I'm sure Adam does this, too, each of us are thinking what agents are we going to use to do what we need to do every day. More than 80% of our engineers use a spectrum of AI assisted to agent it to complete orchestration to build the software that they do every day. And it's not just about writing code faster as we have folks like Adam and I and others who build software, there's a lot more after the software is written, which in terms of pull request reviews. That's what you do to merge the code and validation from a quality standpoint, ensuring that there are no vulnerabilities in the code, doing performance testing. All of these are important in how software gets written and shipped, right? People focus too much on how the code is written, right, in this day of age. So concretely, we are using AI to shrink code review cycles. We are using AI to add unit testing and functional testing so that we find more issues before the customers find those issues. That helps us to ship product with fewer regressions and fewer problems. And the amount of optimization or the amount of efficiency that we are getting out of all of this. I don't think we have a single number today because it's still early in the stage. So we're still yet to arrive at that stable state, but every signal points to us getting a lot more out of that. We are not only using that internally. Adam and I talked about Q2 Code, which is a product that we ship to our end customers, developers and the banks or our partners can use to build on top of the platform. We built that using AI, right? AI tools being built using AI. And so the other sort of big shift in how we think about it is if you heard about the 2-pizza teams. Now the teams are even much smaller because orchestrating the AI agents to write code all the way from thinking about requirements to writing the code to testing is done by half the size of the team. And so that is also a big shift of how we are thinking about removing these transfer of information in the old days versus an AI Pod approach to go deliver that product, right? So a lot of exciting work happening both internally and how we are shipping these products. We are shipping code in production with Q2 Code and Q2 Assistant that we have early adopters. So it is not, like I said, proof of concepts. This is built into how we do work today and how we are shipping products. Adam, would you add anything?

Yes. Just one thing. So I had a chance to interact with a customer that had been using Q2 Assistant, which is our back office-facing agent that helps our customers use the back-office tools in our product. more efficiently. And they identified 3 or 4 tasks around responding to secure messages, around resetting passwords, around managing entitlements where they were seeing substantial, I mean, greater than 50% increases in the efficiency of performing the task. That is a substantial improvement. So we're very early with that. But the fascinating thing for me is people will try anything with that text box that you type to the agent in. And so the great thing is we're also finding out what they want to do, right, organically instead of them telling us by just watching. So the dynamic of the interaction with the agent because people have different expectations than a traditional piece of software is really changing the way we do product management and product engineering in a way that's interesting. And I'm excited to see some of the outcome. And we know that if we can reduce the amount of time it takes to manage the platform, that's time that a financial institution can invest back into selling products, taking care of customers, securing their enterprise. That's clear value that we can get revenue for and that makes us more valuable to our customer.

Makes a ton of sense. And Adam, as we -- a lot of questions we get from investors around like what does the future state Agentic world look like where they're running workflows, making decisions, executing transactions. Where does Q2 sit in that deployment of Agentic workflows within banking? Do they sit above our platform? Are they through it? Does that distinction even matter? What are your thoughts on that?

Yes. I think it's early, but here's the way I see it shaking out. The fundamental problem to some extent, with an Agentic world is that you have to trust the agent in order to allow it to behave autonomously. And if you're not willing to give the agent some level of autonomy, the agent is not nearly as useful because you're just directing the task, right? I don't want to marionette where I have to grab the crossbar and pull the strings all the time. I want a homunculus where I give it a task and it goes up to perform the task and then it comes back and tells me that it's done. So if you set aside old world literary analogies for a moment, when you think about this agentically, I want to consume the services of an agent from a place that I trust. and I trust my financial institution. And so given a choice in consuming Agentic technology around my finances with access to my bank account and my key financial data, it feels natural that an end user would strongly prefer to get one that sits behind the same guarantees that they get from their bank or credit union. And so when you talk about observability, explainability, repeatability and guarantee, LLM technology and AI, part of its value is that it is stochastic. It is nondeterministic. It's part of the charm of it. Part of his drawback is that it's stochastic and semi random. And so I want to know who I'm going to call. I want to know where the building is full of people that will help me if the agent doesn't do what I need the agent to do. So I think our financial institutions have an extraordinary advantage in saying, why don't you let us bring Agentic technology to you behind a TrustedLogin in a secured location. Is this the way that the world will shake out? If I knew the answer to that, then I guess I retire and just prognosticate on things. But it seems very reasonable that we can work with our financial institution and construct a message within the adoption of Agentic technology around finance is tied to the financial institution behind their login and associated with their brand and that trust. Because when I look at what's missing in all of this, whether it's with OpenClaw or Cursor or some of the other things, CloudBook, whatever they rename that thing to, it's trust, right? It's already difficult when you just go on the Internet to trust the images that you see, the text that's being written, the dialogue that happens. And so I think for people to adopt Agentic technology and finance, it's very attractive for them to adopt it through their financial institution because it helps solve the trust issue in a substantive way. That's where we see things sitting today.

That's great. Hima, any thoughts from your end?

Yes. It is -- my perspective is agents are going to be so permeated throughout the organization. To Adam's point, how do you -- how do they operate? I think you have to start naming agents soon. How do they operate within these trusted environments, whether they are coding agents, whether they are CSR or customer service agents or whether they are fraud agents, all of those agents have to operate in a trusted environment, merging this nondeterministic way the LLMs operate and merging that to the guardrails we put on it. And the guardrails we put on it are coming at major areas, right? That's where the context that we talked about. The context can come in through the workflows. The context can also come through how we deploy the LLMs. Maybe Adam can touch on it later. We add a lot of system context to what we want the LLMs to do when the agents are running, and it could be the coding agent or could be the CSR agent. We're adding a lot of determinism so that when the bank CIO says, "Let's hope it works" is not acceptable, right? We want to make sure it works and it works as expected through quality, through evals on validating that we need to ensure when we test it in our environment, it performs as it is expected. So when these agents are deployed within a bank, it automatically gets all the security controls, the identity framework I talked about, all the compliance validation, transparency of decision-making, audit trails and the infrastructure that comes with it, right, the scale because what everyone -- a lot of folks forget is scale is a big part of how -- what we manage, right, with 27 million end users and growing every day, our infrastructure today scales. And then as agents are doing the work of some of the back office or agents are coming in to talk to the digital banking APIs like Adam talked about, it has to scale based on the load that's coming from these agents. So all of this has to come in together from a technology standpoint that enables agents to run within the infrastructure around the infrastructure calling into our APIs or the digital banking or the fraud or the pricing APIs. The value is the bank's engineering team don't have to build any of this, right? That comes from the existing infrastructure, whether it's running through our latest incarnation in the cloud that we have finished over the last couple of years. But then the agent and the bank can focus on deciding what is the logic, what is the context that the agent should work on, what are the conditions that they want to put. And we do know that not all banks, financial institutions, credit unions want to operate with the same context, right? So one of the things about it is it's not just one agent that has one shape, right? We've got to give the capability to these financial institutions to create the right agent that works for them, that is unique to their financial institutions operating controls. We handle everything underneath that definition of the right policies, the right context so that they can focus on their outcomes. And so we're pretty much taking what we've been doing and expanding that infrastructure to the Agentic layer. That's how I think about our role in deploying these -- how the financial institutions are deploying agents.

That's great insight from both of you. I want to shift gears a little bit and just talk about kind of Q2 as a player in financial services relative to other names that we see in the space. So Hima, I'll start with you on this one where a question that kind of comes up consistently is what keeps either a larger tech company or even a smaller one from entering this space and kind of displacing us or challenging our right to win? What gives us the right to continue to win and why might that be durable?

When someone says, why can't a small company, and I think I got this from Adam, why can't a small start-up do what rewrite digital banking? We can do the same. We do it all the time. When we think about new features, new products, when we have to redesign something, clear up the tech debt, we do that, right? We decide to rewrite the code or we redesign an existing code or fix it. What -- where that code runs, I think that is where I see the advantage. I'll give you an example of what happened 6 weeks back when a complete region in AWS failed. And we were probably one of the very few enterprise software ISVs that AWS has that were able to be up and running because we do have very strong active, active multi-region architecture. That's an investment we make. That is an investment we continue to look at from an operational standpoint to ensure that when we are deploying software, when we are testing software, it works uniformly across multiple regions and it scales, right? That is something. It takes a lot of investment. It takes a lot of collaboration with our partners like Cloudflare and AWS so that those environments are scaling and protected. In addition, over the last 25 years or so, we have been subjecting this environment to regulatory pressure, FFIEC testing, OCC, FDIC, any regulator that looks at our stack and goes through all the validation that we are complying with that, right? So that's something that not -- that doesn't happen in a matter of a day or a year or so, 25 years. We've been continuing to strengthen our posture and positioning and that infrastructure is what we offer compared to a nimble startup that is building the software. As we talked about in the beginning of the conversation, building the software is probably 10% to 20% of what the end-to-end value is. It is the complete operational environment is what is hard to put up. Incident response, right? A lot of the organization within the engineering and the cloud team are on call 24/7, so that if something happens, we have broad enough coverage globally to ensure that any time of the day, night a problem happens, we are responding to it.

Adam, any thoughts from you?

Yes. I think about it sort of economically, right? If it takes 2 things to make a product and one of them just got a whole lot cheaper, it means the value of the other thing just went way up. So if you can generate code much less expensively, that's great. And if everyone can generate code much less expensively, then that means design, discretion, operations, observability, architecture, all those things are now more valuable because the other thing is substantially less scarce. Like I saw a Lakers game the other day. If you went and you dropped the hoops from 10 feet to 7 feet, I could dunk. That doesn't mean I can go out there and compete in pro basketball because the hoops just got dropped for everybody. So there's a lot more to it than just generate 100 lines of code, generate 1,000 lines of code. There's even a lot more to it than the coding and the deployment or the construction of products. So I think it's fantastic, these AI changes. I'm kind of grateful that the AI revolution came for engineering and coding first because I think Q2 is really well poised to take advantage of it. I think that it's going to move to other parts of the business and other parts of the industry. But economically, as [indiscernible] that really are the transformation of information from one symbol set to another set become commoditized and automated, every other part of the process becomes more valuable. And so the question is not like do you have access to this thing? Like we all get access to the tools. You just -- you download it, you pay for your tokens, you do your work. The question is, from a competitive perspective, do you have a culture? Do you have the hunger to actually continue to progress? Can you attract the kind of talent, some of which, frankly, is entry-level younger talent that can take advantage of these tools. That will be the difference. And then there's something -- and it's not real quantifiable, but I'm going to put it out here anyway. If you don't know what kind of company you are and what your value is and what your mission is, when you take away this thing that used to be very difficult and you make it easy, you don't -- you really don't know what kind of company you are. I think the companies that have a good sense of what their value is, who their customers are, what the mission is, what the culture is, those are the ones that will come through the AI transition the most effectively because they have an embedded capability for continuous learning, continuous improvement and a cultural hunger for getting better. Lacking those things, I think AI is a tough thing for a company to adopt.

And maybe one final kind of follow-on from a competitive standpoint. As we think about some of the names that we've been competing with for the last 20 years and these kind of established banking infrastructure players, in an AI-driven world, how does Q2 look to stay ahead against them specifically?

Yes. So part of it is talent becomes more important. Taste and discretion become more important. And I don't need to name names, but go look at legacy infrastructure players. Go look at the big payments companies that are also in digital banking technology. Go look at some of the subscale entrants that are struggling to put together enough revenue to get some level of accretiveness in their growth path. They are all in a very challenged space, in my opinion. We have challenges as well. We have challenges culturally because we've got to get people to understand their value is not the thing that they used to do that the AI can do now, whatever it is, writing code, closing books, putting in journal entries, evaluating source code, doing code reviews, that wasn't really ever the real value of the job. The value is their ownership of the end-to-end delivery of value to the customer. And as it gets easier to perform the [indiscernible] automatable tasks, their ability to be imaginative, their ability to think differently. I used to have the crutch when I didn't want to think very hard about how I would deliver a feature to a customer to say, "What's going to be really hard to code it that way?" So we're going to code it this other way. The customer would say, okay, well, that makes sense. Now coding things is not the challenge anymore, right? The challenge is how much can you imagine? How much discretion can you apply? How much domain expertise do you have about what the right way to solve a business problem is. And so that's really exciting, but it's also kind of challenging because in the absence of the constraint, many of the ways that we limited ourselves in the way we drove the business and delivered value have gone away. The more of that we can tear down and the more of that we can pursue, it's very exciting. And so I was talking to our Chief Legal Counsel, right? And I said, during that time, I asked you how many of our contracts had language like this? And he said, I don't know, but some of them, and I don't know which ones. That's now a 12-minute prompt and probably a couple of thousand tokens, whatever, to solve that answer. And so you can ask a question and get an answer more rapidly. You can get better data for better decision-making. And so we can strip away a lot of this ad hoc, rule of thumb kind of culturally embedded decisions, and we can replace them with real interrogation of data that previously would have been just too expensive to do. I think Q2 is in a better position to do that because of our culture and the way we interact with our customers than the legacy, the infrastructure incumbents. And I think to some extent, even if you look out at new market entrants, whether they be massive or tiny, their need to acquire the domain expertise around the space because banking is a simple business, I think hamstrings them to some extent. And I'm not saying they can't make progress. I'm just saying that the addition of AI to us means we can continue to outrun them.

Hima, any thoughts from your end?

No, I think to reiterate what Adam has said, it's something which -- I think the market is sort of changing so fast. It's hard for every financial institution to keep up with all the changes. They're looking for a partner who's enabling them to go solve that versus having to run with all the changes. And that's something which is the value we offer going back to the talent and the approach we have within the organization. And so that's something which we are enabling them and helping them. So, yes.

Yes. I think maybe that kind of brings me to my last question, too, which is from investors, we get talking about like the sectors and markets that we serve, specifically as a software company and they've been curious about this. So banks may have been viewed as being slower to adopt new technology in prior cycles and regulatory compliance, legal risk, board scrutiny, all the things that we've been talking about. Maybe Hima, I'll start with you. How do you actually move customers forward on that AI adoption? And what does SaaS actually mean to banks and credit unions?

I have an interesting perspective given I've been here 2.5 years and came in when we were kicking off the cloud transformation and the kind of conversations I would have in terms of a little bit of friction, a little bit of concern in terms of putting that particular financial institution in the cloud earlier than the others. And so it's interesting 2.5 years later, what I'm seeing in the market, it is very different. They're not resisting it. They're asking for it. I'll give you an example. We are shipping a new fraud product called user activity monitoring. And it has -- it's based on a bunch of detectors that tell if the user who's claiming to be a valid user operating within digital banking and working with the financial institution, they come back and say, "hey, can you add this new detector that is using volume of a transaction, right? We may not have had it. And they're willing to work with us, knowing that we use ML and LLMs to go build those detectors to quickly add a detector for that particular signal and push it back to them, right? I wouldn't have had this conversation 2.5 years back when they would be worried about how we are able to do that. And so with all my conversation, it should not -- it is no longer should we do AI, right? It is about how do we do this in a very safe, compliant and doesn't expose them -- the biggest things the financial institutions are worried about is exposing themselves to customer trust, customer experience and regulatory risk, right? At the end of the day, we want to make sure that customer experience, customer trust, along with the regulatory risk doesn't take a next level. And so with those guardrails, they're saying, I want to move, what's the framework to move that. As we all are, Matt and our conversations are all about what are we doing with AI, how are we using AI to help our customers and their end customers solve better problems and how do we do that safely. I think the leadership, the CIOs, the business leaders, other banks and the financial institutions are also being asked the same questions by their leadership, by their Board. So I think the market is a little different right now given this is a big tectonic shift that's happening in the industry, right? It is like the Internet wave of the '90s where it's not optional. It is about I want to do it, how do I do this with a trusted partner who I'm comfortable with, who gives me that tools that continue to operate in the same environment. And so they want to work with someone who understand all the risk constraints, who understand all the market requirements from our regulations, compliance and who knows how to scale infrastructure, right? That's something which is continues to be important in terms of what they're looking for versus don't make me the first user of AI, like Adam said, right, some of the early feedback that we're getting from our early adopters on the AI Q2 Assistant is incredible. They're seeing the value that comes from it. And so I think they sort of slowly moving from the exploratory stage to now thinking about, okay, how can I get ahead of the other banks? That is one of the other things, other financial institutions, banks or credit unions. That's one of the things they are also realizing is how do I make this a competitive differentiation versus trying to use that as a defensive strategy. So that changes the conversation we are having because that puts a sort of a time constraint or urgency from a timing standpoint. We're going to do this. How can we do this in 90 days? That's the kind of conversations we're having with early adopters versus come back and talk to me in 2 years, right? That's a fundamental shift I'm seeing because I think a lot of it is driven by their own work that they're doing outside of digital banking to leverage AI because AI is pretty permeating across every part of the organization, and they're seeing the value of it versus maybe cloud was a different story, right? And so that is sort of giving us that more aggressive partnership with these financial institutions to go deploy Agentic solutions in their back office, in their coding environments. And so I still think of this happening in different cohorts, like any other adoption curve, we are in the first 20%. I would have thought the first 20% would have been a longer time spectrum, but that is the difference that I'm seeing now, which is that first 20% wants to do it now, wants to do the number of early financial institutions signing up for early adopters is pretty high. I wouldn't have expected this even 6 months back when we were thinking of these new products that are Agentic in nature, whether it's Q2 Assistant or our fraud products or Q2 Code. I have a feeling that next 60% as they start seeing their peers in other industries, and we do this a lot, right, because we have a large customer base. We're always talking to them in our QBR, CBRs, giving them updates on how their peers are doing in other markets. They're going to come pretty quickly and say, okay, I want to go deploy that. And that's something which will be a good problem to have for us because our passion comes from seeing a lot of these products being used a lot in the larger market. And so I would segment the market as the first 20%, and then we'll go to the next 60%, and it will happen pretty quickly as early adopters succeed. But I feel like we are in the cusp of late early adopters, early next 60%. Adam, that's what I'm seeing. What are you seeing?

Yes. I'm getting questions through folks I talk to banks that are coming from their Board, like they said, my Board is wearing me out about what are we doing about AI? What are we doing about stablecoin? What are we doing about tokenized deposit? What are we doing about quantum computing? These are Board-level conversations now. 3, 5 years ago, the boards were not asking about technology topic. One of the advantages of the enormous hype cycle around AI has been -- that it provides you with air cover to talk about why it's important. It's in the news every single day. I mean it's an extraordinary part of the news cycle that has not moved or shifted. And so we know that our banks and credit unions are generally on the acquiring side of merger and acquisition, partially because we have a great product, it makes it easy for them to do so and partially because we tend to win the customers that are more aggressive and they want to grow and they want to use technology to do it. But when somebody says to me, I was talking to a Board member and they want to know if you guys can provide us an MCP interface to the back end, so we can build our own AI tooling against your platform, that's not a conversation I anticipated in 2026. But I'm super excited that's happening because I love to work on those things and talk about that stuff. So for 20 years, at least that I've been at Q2, this story has been about we can give you the technology that makes you competitive with the larger bank and whether they're a little bigger than you or way bigger than you. We can take away some of the disadvantages you feel like you have, so you can lean on your advantages about being in the community, being specialized, being in a niche. That has become more true and not less true. When the mobile phone came out, I thought it was one of the most fantastic things ever because I don't know if you noticed, but Bank of America and Bank of Waxahachie, they get the same screen size. They're the same on the mobile phone. There is no reason that the experience can't be as compelling. AI, I think, further accelerates that trend, right? There is no reason for a small financial institution, I'm talking about $2 billion, $5 billion, $10 billion, not to have the same level of technology as a $250 billion super regional or even approaching the same technology level as [ Chase ]. And I might argue, in a post-AI world, because of their size and some of their nimbleness, I think you could see some of these smaller financial institutions using technology to really carve back some market share from some of the largest banks or some of the big mid-tier banks because they can move faster because they're not hung up by committee and because they -- to some extent, they don't compete with each other nearly as much as they compete with the folks outside of our customer base. So it puts us in a really interesting position to be able to work with them on this stuff. And man, I'll tell you, they are hungry for it. We have -- I have conversations today on every one of these topics with a $2 billion bank in from East Texas. And so it's top of mind.

Awesome. Well, I appreciate both your time today, and I think investors and analysts are going to find this content really valuable. Before we close out, maybe I'll give you both one more chance to share any final thoughts that you have. And Hima, I'll start with you.

I want to leave with what Adam said, right? This is AI as a technology is enabling everyone to be a technologist. It's enabling everyone to move as fast as someone with a lot of investment. And our job at Q2 is to give -- help them move as fast as or even faster than the person sitting next to them or an institution that is even larger than them. I'm super excited about the journey that not only we are taking, but what we can enable our partners, customers. So it's great times to be here.

Yes. We're really privileged at Q2 to have the roster of customers we have who are so critical in building our business, in trusting us and in providing us with kind of a North Star for executing the mission. AI as an enabling factor to allow us to go faster for them, and it's just tremendously exciting. I tell people all the time, when I was 9 years old, if you'd ask me what I wanted to be when I grew up, I probably would not have said I'd like to be the CTO of roughly trending towards $1 billion revenue software company that works with banks to try and make community banking really great for everybody in the U.S. And looking back on it now, I cannot imagine having chosen anything that would be more fulfilling. And so as each of these technology waves has come, the Internet, mobile, microservices, cloud, AI and AI probably is as big as all the rest of them combined. It's just an extraordinary opportunity to reinvent and rethink the way you apply understanding business problems, creating value through solving those business problems and then working with people that you actually care about to try and do something meaningful. And I think that resonates with us and our customer base. And I think that's where we get a lot of our loyalty, and it's where we get a lot of our forward momentum in the space.

Awesome. Well, with that, we'll go ahead and close it out. So once again, thank you both for the time. I really appreciate you both taking time today. Thanks.

Thanks, Josh.

Thanks, Josh. Thanks.