Hello everyone, welcome to JP Morgan Boston TMC Conference. My name is Alexei Gogolev, Head of Vertical SaaS Research. Today I'm delighted to be hosting Intapp Management. We've got John Hall, company CEO, and David Morton, company CFO. John, David, welcome. If I may begin the conversation, framing the story and kind of discussing what to watch out for. First of all, there is a debate obviously whether AI reduces demand for professional services. John, what are you seeing that suggests AI is actually a tailwind for your customers? Things like project matters, faster turnarounds and do you agree that it's actually less of a headwind for your business?
Thanks for having us, Alexei. Intapp focuses on the large financial professional services firms. We were a bootstrapped company focused on this specialized, highly regulated end market for many years, and we've built the company through multiple technology generations. Through that time, these firms have grown. The industries that we serve, the law firms, the accounting firms, the consulting firms, the investment banks, private capital investors, real assets investors, basically the large partnership segment of the economy, which is about 3% of the global economy. It historically was very overlooked as a sector, but is quite significant and it sits at the center of capitalism. These people do incredibly important things to facilitate fundraisings and financings and the allocation of capital, M&A, on the other side, litigations and restructurings in difficult times in the cycle. They are central to the way that capitalism actually functions.
What we have experienced as we've grown the business is that as the economy develops nationally and globally, these folks win more business. There is more natural demand for their services regardless of how they're fulfilled. I think what people are kind of focusing on is the efficiency of delivery. From a demand standpoint, I actually think the demand is a structural component of the way that capitalism functions because the role that these people play are in facilitating all of those types of transactions and dispute opportunities. They may go through some very interesting pricing model changes. They may move from hourly to fixed fee in some parts of the market, or they may charge in other ways.
The actual fundamental requirement that you have these folks participating in their highly regulated industry to serve the way that capitalism happens, capital flows happen, and all those other things, I don't think it's going away. Our experience has been, particularly in the enterprise firms, which is where we focus, the top 2,000 firms have about 70% of the TAM for us. That group is actually getting larger. The top 10 private equity firms now raise something like 40% of all private equity raised. The top law firms have gotten bigger over the time that we've worked with them. There's economy of scale. There's also demand from the larger national and global clients to have a better footprint to serve them.
In the accounting industry, very interestingly, the private equity community has come in now, and they're driving a huge consolidation trend under the Big Four of the next 100 firms to try to build up a more and more capable multinational set of businesses. We've oriented our whole business towards the enterprise segment because that group is growing much faster than these industries as a whole, and that's where the technology demand and opportunity is to support the growth there. In good times and bad, even through 2008, 2009, we grew the company and bootstrapped to these firms. We're experiencing that now as well.
John, as you execute towards a rather robust target of $1 billion ARR by fiscal 2029, what are the most important indicators that you would highlight to us? You've talked a bit about demand, so pipeline and bookings, but also maybe some product and AI solutions, maybe some NRR trajectory that you think are worth highlighting.
Fundamentally, the company is growing. David can talk to you about some of the numbers, but we have a very strong land and expand model for these firms. We have a good footprint, 2,500 or so clients, a large relationship with many of them, but a lot of expansion inside. Both a land motion and an NRR expansion inside these firms across a range of offerings on our platform, helping with the firms with their go-to-market, their business development activity, helping with their compliance activity. This is a core part of our story and historic value proposition is serving the highly regulated nature of these firms. And then a whole set of things around their efficiency and use of resources in time management and resource management.
The, the way that we are bringing AI into the story is kind of what you would expect in all the business processes that we have been supporting over the years. We have real depth of expertise, and we're bringing a agentic story now into each of those areas to help these firms as businesses at scale get the benefits of the agentic opportunity in executing the way that their firm operates. It's a little bit different from some of the other stories that you're looking at in the market, where folks are really focused on the practitioners in the practice areas, which is also by itself a huge opportunity. The business of the firm is historically our footprint, and strategically is a huge opportunity to bring agentic capabilities in, particularly in a compliant way, which I think is a real differentiation for this market.
We can talk some more about that.
Well, you spoke a lot about agentic reality, so if I may ask about Celeste. When you say Celeste is agentic, so not just GenAI embedded, what are the key end-to-end workflows you expect agents to run in production? What makes those workflows uniquely suited to Intapp versus some of the horizontal automation?
Right. The founders of Intapp are still with the company and leading the company. This is not a company that's in the hands of a next generation executive team. The folks who built these products and met these clients and introduced new technology to them are running the company today, and Celeste is our next offering. It is a from-the-ground-up, AI native, agentic platform built with compliance in mind. It's built specifically for the compliance requirements of these highly specialized end markets. The management of conflicts of interest, independence, material non-public information. We talk about it as professional compliance. It's the issues that really are unique to these high-end professionals and financial services firms.
The Celeste architecture was designed in collaboration with our clients, which is how we built the company all these years, working directly with both the IT community, but the risk and compliance community inside these firms to build an architecture that can provide the protections and guarantees for the firm's client information and market-sensitive information that is very different architecturally from some of the other offerings that you might see from some of the horizontal players that are bringing agents to market. The simple way to think about this is you have a practitioner in one of these firms who wants to do some AI analysis of documents, and most of the first generation tools are encouraging them to drag documents out of governed systems and put them in this AI and start to ask it questions.
What the firms are discovering now after trials for a while, is that they are accumulating a new repository of ungoverned information that is commingling or at risk of commingling all their client information and answering questions about it and encouraging people to share those questions with other parts of the firm, potentially in contravention to the compliance requirements, the ethical walls, information barriers that need to be in place for professional responsibility and regulatory reasons. The Celeste architecture is designed differently. We begin with an understanding from our history in the compliance business for these firms that you have to do information governance with an understanding of professional compliance. It's a totally different design, co-designed with the firm's compliance departments in addition to their AI and IT departments.
What it enables you to do is to build agents and a whole agentic strategy that gives you the benefits of AI automation, all the things that everybody is excited about, while respecting from the ground up the firm's professional obligations that it's signed up to with its clients and its regulators and the courts. From this perspective, there is a lot of excitement because firms have had this real tension with the first generation of AI tools. As exciting as they are, how do they protect their traditional and required confidentiality management and information management obligations? The application space is across the whole firm. You could be looking at business development activities. You could be looking at compliance activities themselves. You could be looking at the revenue activities of the professionals.
You have a compliance framework that enables you to build that automation in a way that doesn't violate the firm's principles, and it's a huge differentiator.
John, any proof points that Celeste is actually working? Do you have some data points like reduction in human touches or task completion rates, daily active usage? Like, what's something you can share?
Yeah. We launched Celeste this quarter in limited availability. We announced it at our Amplify event in February, about 2/3 of the way through this quarter that we're talking about, Q3, our fiscal Q3. We had very positive responses. We had several launch announcements with that event, including several Celeste clients that were speaking, Starwood Capital was one of them, about the success stories they've had. Over time, we will bring it out in general availability. We're working with each of our clients in each of our segments today to make sure that we have enough reps as an organization to be able to show client success with each and every rollout when we scale it up. That's the phase we are in today.
Some very exciting results from some of these limited availability adopters, where they're looking at their labor budgets for some of these programs, like the way that they vet incoming business in their conflicts or compliance department, and can plan for agents to take over portions of that work, a meaningful portion of the work. If you look at the P&L of these firms, there is 20%-25% of the budget today that goes to a lot of these support functions, not the working professionals in each of the practice areas or each of the investment teams, but the actual supporting team where there's a gigantic opportunity to support with automation.
You look at the portion of the professional's time, particularly as you move up to the more senior ranks where they start to shift from working on client work to working on managerial responsibilities for the firm or for their practice area or strategy team and all the compliance and all of the leadership functions that go along with that. There's a pool of time that those people spend that could be 40%, 50% of their time that we think the agentic opportunity on the business side of the firm could make a profound difference in how those folks work. Our engagement today is across each of those functional areas and at both of those levels.
Some of our earliest users actually, the example we gave at our launch event was that the CEO of the firm is interacting with Celeste and getting questions that they previously would have had to ask a person to go and find information and bring it back to them. A lot of what we're doing is bringing the power of AI in this compliant environment to some of the most senior people who previously were intermediated to their business information through humans. I think just the opportunity here on the business side of the firm is gigantic. A lot of what we've put together working with the firms is about moving into those spaces here as we go to general availability.
Okay, that's actually pretty fascinating. You're basically saying you're going beyond your typical IT and software budgets into broader operational value of those firms.
Yeah, there's really three spaces where we can pursue budget. There's the traditional IT budget, where we have a long-standing relationship with the CIOs. Absolutely, there is spend and AI spend coming from there. Many of these firms have created a new AI budget specifically. They know that this is coming and they need to stay competitive. There's often a new AI leader that's been brought in who's running the program of evaluating all the first-generation tools and trying to figure out where the best opportunities are going to be and what the issues are going to be. Those groups are particularly exciting to talk to because they have a little bit of experience now with the compliance challenges that come from the first-generation tools.
There's been a couple years of folks getting some experience with the issues, and when we bring them the Celeste architecture, they say, "Oh, you have addressed a lot of the things that I was experiencing in my previous attempts here over the past 18 months." The third bucket is this labor pool. That's not traditionally an area that we have pursued, but the opportunity to bring agents in to take over for hiring in certain areas of the support functions in particular is hugely attractive to these firms.
One of the other things about the labor opportunity too is, in addition to comparing our value to what it would cost to hire someone, there are things that the agents can do working around the clock and searching the whole internet to find information to vet for due diligence or continuously monitoring a particular client, not just at the point of bringing it on board, but throughout its life cycle, that no human could have done. There is a quality and upside angle to the value proposition in a lot of what we're doing that really is compelling to people almost on top of the cost savings story that a lot of people talk about with the agentic pitch, which we definitely have.
I really think one of the most interesting things is how powerfully you can accelerate the firm's ability to manage its quality and output for clients and its risk management program quality at a very reasonable price.
John, one of the key themes that has been coming up a lot at the conference, at all of my meetings, has been this governed AI notion. Can you talk about what's compelling about it? What are the key concrete technical and process controls that create this defensibility? How do you prove to a CRO that it's working? Maybe talk about some walls enforcement, some data provenance, model rerouting. How can you make sure that consumption is in this governed way?
Right. We use this phrase too, governed AI, and when we talk about governance, we're talking about the professional responsibility compliance of the professionals in these firms that we talked about earlier. One of the most important things to understand about this set of industries that is very different from any other normal business is that for every investor that they work with or every client that they take on, they custom negotiate an engagement letter or engagement terms for the client relationship overall and for each engagement or matter thereafter. In those letters, they are making commitments to how the deal will be staffed, who will be able to have access to the non-public information associated with that deal.
They make commitments about what level of person, each client is very interested in what other clients does the firm have, what other information does the firm have. You also end up making a list of people that you have to keep away from your deal and your engagement team. Then as that team changes over time, if you have a long-running engagement or a long-running deal or a long-running matter, the people at the firm who are working on that are not the same at the end of the engagement as they were at the beginning. There is a rotation of humans through these projects.
The information management associated with this simple fact, and these rules follow from the duty of loyalty and the duty of independence and all these professional obligations that people have, when that translates into procedure, the information richness of the body of promises and negotiations that have been made with each and every client and each and every matter over the course of ten years back, and then every deal that comes in has to be compared and contrasted against all of those obligations. This is the information set, the proprietary information set, that Intapp has historically managed for these firms. We are very unusual in having command and an infrastructure that really understands this. It is truly industry-specific, and it is directly tied to the compliance obligations of the firm. It is an existential issue if you leak your client's information in the wrong way.
It's not just a slap on the wrist. The reputation of the firm is on the line, so they take this extraordinarily seriously. We have always been the people that are trusted to manage that incredibly bespoke universe of client obligations. If you have, you know, 100,000 clients running, you know, 10 or 20 engagements or matters a year, it's a gigantic number of overlapping requirements. The whole design of Celeste is to ensure that you can deploy agents into that environment, and it will understand and respect. Each agent will understand and respect that detailed universe of expectations about who's allowed to know what, including the agents, is very difficult from an information governance and design point of view to keep all of that running successfully.
We are leveraging our years of study and history to become the experts in this compliance area in our, in our design for Celeste. Regardless of the business process, you could be working on client research, you could be working on client due diligence, you could be working on the financials of the client, you could be working on, you know, your possession of material non-public information on the part of a deal. All of that has to have this governance framework in place. It's just a real trick. How is AI gonna work for these firms? They have such a unique compliance set of obligations that there's a special design that you have to put in place. If you can do it, then they can benefit from the AI revolution in the same way that some of the more traditional companies can.
There's a lot of excitement and a lot of interest, but there's also a lot of design that has to go into being able to do that successfully.
John, how about, the fact that Intapp sits on this highly sensitive firm data? How are you balancing more context improves AI notion that obviously a lot of firms do with data minimization and client confidentiality? obviously, you know, when you think about training AI on customer data to benefit other customers in the industry, like, how do you draw this line?
This is so central. The encouragement of all the AI systems is to drag everything in there. As exciting as that may be, it is a terrible risk to the firm's compliance posture. The reality is each firm is making a trade-off in their risk management posture. How much do they want to commingle information in order to get the potential benefits of commingling it for the purposes of AI analysis? How much risk are they introducing to each client? This is a policy decision that a firm risk management committee and AI committee have to make together. One of the things that we're finding is our experience in this and information governance in the earlier generations, which had analogous issues.
The document management generation had the same set of theoretical issues, where were you gonna put everything together into a document management system, or were you going to segregate it out? How did you balance the needs of knowledge management versus risk and compliance management and client confidentiality management? We have a lot of history and language and a grammar and a whole infrastructure technologically to help people from a policy standpoint capture all these commitments and measure those and weigh those, and Celeste inherits all of that. We're doing the same maneuver, but now with the agentic opportunity. I actually think that the benefits are enormous, but it's not going to be the same as you see in one of the corporate environments where you can just throw everything in there and, you know, it's not as big an issue what it's talking about.
At the same time, the value of the work that these professionals are doing is so enormous to the facilitation of deals that I think there's a lot of value that you can unlock with this design and still protect the professional obligations of the firm.
John, how about Walls for AI? It seems like it's becoming a standalone platform capability, and it does expand your reach beyond your own app. Maybe talk to us about how do you monetize that without cannibalizing existing module pricing. I'm talking about covering Copilot and then your partnership with Claude, your partnership with Harvey.
Right. This is an interesting and exciting opportunity in addition to Celeste itself. This is taking our compliance and governance capabilities and allowing firms to extend those into some of these other AI-based generation tools. Walls for AI is a governed AI offering that if firms have our platform, they can extend its governance protections into many of the AI technologies that they may also want to work with. We have always been a centerpiece company, a platform company for the firm with a wide ecosystem of technology and services partners to serve them. We have a open architecture that encourages people to look at Intapp as the centerpiece of firm governance and operational management on the business side of the firm and then extend to many of these other data sources and data repositories.
Walls for AI continues in that tradition, and there's incredible demand, as you can expect, from folks who need to manage many of these first generation tools to get the better value out of them.
Moving to your partnership with Microsoft, more than half of your latest deals have involved Microsoft sales teams. What is structurally different about Microsoft influence deals, and how repeatable is things like Azure investment dollars as an accelerator?
We have a longstanding partnership with Microsoft mainly because these firms in these markets tend to be very Microsoft oriented. If you think about lawyers in Word and consultants in PowerPoint and bankers in Excel, not to oversimplify, but you know, people are in Office 365 all day long. Microsoft has an incredible position not just in the front office there, but in the whole back office infrastructure with the CIOs that support it. This was a big input to us as we built the company, was that Microsoft was a central pillar of the way that these firms had chosen their IT architecture strategy. We have a longstanding relationship there. There's a technology partnership. We worked with the head of Copilot as Walls for AI was being developed, for example. There's a sales and marketing go-to-market partnership.
Several of the aspects of that have enabled us to co-sell with Microsoft into many of these firms. Take advantage if the firm has a Microsoft Azure spend commitment, these MACC agreements where they sign up to spend a certain minimum amount with Azure every year, which many of these firms do. They can now buy 100% of Intapp's offerings off of the Azure Marketplace under that agreement. In many ways, Microsoft has taken the budget issue off the table for us because the firms are already committed to spending X with them, and they can buy the Intapp platform or expand their Intapp platform through that Microsoft contract. Because of that, the Microsoft sales team is highly incentivized to collaborate with us because they get full quota relief for helping us deploy our software there.
Our sales team and the Microsoft sales team work very closely together on these accounts because the interests are aligned. We're very happy about the long-standing Microsoft partnership, and the CIOs and the firms are too because they really need our platform to interoperate with their whole Microsoft investment. That's been a big part of our story. As we've moved more into the enterprise where more of these firms have larger MACC agreements in place and larger commitments, that's helped us with the enterprise both land and expand motions with the Microsoft firms.
David, maybe this is a more of a question for you. As John was saying, as you go more into this enterprise and co-sell with Microsoft, do you expect ARR to become a bit lumpier because of things like stage conversion and partner source mix?
I think we've done a good job in balancing both the linearity within a quarter as well as within the year. Traditionally, our biggest renewals will be around the December and June quarters. With that said, with our rate of pace of innovation, what we continue to deliver to our end clients, we see that aptitude continue to grow. You've seen that accordingly in our results. As we think about how we continue to scale to $1 billion and beyond, we feel like we've got a great coverage model.
You know, especially going into next fiscal year that will continue on with that growth.
Okay. Perfect. John, back to you. With those multiple AI partners that we just discussed, how do you maintain a model-agnostic posture in practice and while obviously still going deep technically to deliver the best-in-class outcomes?
One of the important design inputs to Celeste as we worked with the CIOs in the market, one of their key requests was, "Please help us maintain flexibility with the models," because the investment pace is so torrid and the pace of model development is so fast. If you just look back at the last 12, 18 months, which model is in the lead has really moved a lot, and I think that's heavily influenced the decision-making of these firms. We built Celeste as a model-agnostic agentic system. You can use OpenAI's models behind it, you can use Copilot or Azure's models behind it, you can use Anthropic's models behind it, that was an important decision.
We also have a Anthropic partnership that we announced in February because obviously there's a lot of attention there right now. A lot of the firms have asked us, "Can you please have that capability?" Anthropic was on stage with us with that announcement in February. We were very excited about that. We're trying to maintain a capability for each of the firms to use their own model choice. Many of them, by the way, have large Copilot contracts. Microsoft was very successful getting them to sign up as part of their minimum Azure spend to have Copilot in there. That's an important component of this market. As much excitement as there is in some other areas, I think, you know, it's important for people to appreciate in this market how strong that footprint is. There are also enterprise ChatGPT agreements.
There are ChatGPT agreements. There are also Anthropic agreements. We allow the firms to use the one that works best for them. At the same time, a lot of the emphasis that we're bringing in is this compliance capability for the agents that are being built independent of which model that you're choosing to work with.
Maybe that's another question for you, David. Seats are now less than 50% of ARR, and Celeste enables this platform plus consumption pricing. How do you expect this mix to evolve over time, and what's the buying center or procurement motion for consumption-based AI for these firms?
Yeah. I would maybe step back and just orient around that we do have the ability to match our clients' requirements for how they view their own payment motion. Clearly, we've got a lot of enterprise arrangements where it's just beyond a seat. DealCloud has traditionally been seat-based. Even as we've started introducing Celeste, yes, clearly there's some material platform fees that get into, you know, fair usage and then consumption on top of that. You know, we've got three or four main ways of how we can specifically orient around our pricing and packaging that help reflect the value that we're delivering to each of these end clients. Longer term, will that evolve? Certainly.
Do I have a depiction of, you know, what I view that as, you know, three to five years from now? Not at this point in time.
Right.
Just know that, you know, Intapp has shown its ability to adapt and be flexible and work with our end clients.
Makes sense. John, maybe question related to that to you. What design choices are you making so that consumption pricing doesn't create some budget anxiety?
Yeah. I think this is one of the other pieces of feedback that we have received from the market and the CIOs in the market. As exciting as the LLM approach is, there is a difference between problems that are well suited for probabilistic answers and problems that might be easily solved with deterministic methodology. There's an opportunity in Celeste agents to both use all of the agentic capabilities that people are talking about, but also to tie it into deterministic workflows for areas that don't necessarily drive token consumption, and you can get a lot better price performance.
I think this has been less of a conversation over the past few years, but I think going forward, the price performance discussion about when is it appropriate to deploy probabilistic techniques and drive token usage to solve the problem or part of the problem, and when can you deploy deterministic methodologies with technology to really get the best answer, correct answer, but also cost-effective answer for the user or for the process, is going to be part of the art of how these things are incorporated into the business process.
I think our history working with multiple generations of technology across a range of areas, and an intimacy with this end market, its compliance issues, causes us to be a little bit more pragmatic about what the design should be to help firms get the best, most accurate answer with the conversational experience or the agentic workflow experience at the right time. I think there's been very positive feedback from firms, particularly as they're starting to see these token bills come in, about how are they going to manage this as the firm's usage scales. That's a key design principle in Celeste, is to help people be able to control that as best they can.
That makes sense. Maybe a final question to you, John? Cloud migration, looks like there's about $100 million still remaining on-prem. What are the top blockers today, and what has changed such that you now expect urgency to accelerate this cloud migration? Are you seeing maybe some organizational inertia or maybe some implementation disruption? What could re-accelerate that cloud migration?
The key for us is that the firms that still have some on-prem from us, that are some of our original customers, also have cloud. It's not that there was a time when the industry was nervous about cloud. That time is long past. Everybody has cloud in some way in their system. It is a pragmatic question, when do they move off of an installed system to a cloud-based system? We've been making great progress. We report numbers on this progress each quarter since our IPO.
The AI opportunity, the agentic opportunity, I think, is causing those last stragglers to say, "Oh, we have to do this now." I think we have a driver for those folks for whom this had not been a top priority pragmatically to say, "I need to move to the cloud so I can take advantage of Celeste and the agentic opportunity that you guys are bringing." I'm very encouraged by the response there, and we're grateful to the clients who have supported us for so long. I'm excited that we can bring them such a new generation of technology, and they're gonna be able to move into this new generation with an AI story for their firm and an experience for their partners that is just mind-blowing.
I'm comfortable with where this is going and the pace that we're gonna be able to get folks there.
Great. Well, thank you very much, John. Thank you, David, for being with us today, and appreciate your time.
All right. Thank you very much, Alexei.