All right, let's kick off day three of Communacopia. I am pleased to be joined by Navan, Ariel Cohen, the Co-Founder and CEO, and Aurélien Nolf, the CFO. Thank you guys for joining us. Thank you for having us. Thank you for having us. Of course. So take us back. Navan was founded in 2015, but in many ways, you guys sound like an AI-native company. So talk to us about your strategy building the key technology building blocks of the company. What early technical decisions paid off? When you think about business travel, it is actually an extremely traditionally manual process. Somebody is sending an email to a travel agent, then the travel agent needs to do a lot of things. Do not think about it as, I am flying to New York, I have one hotel, and I am coming back. It tends to be multi-city. People are changing it a lot. People are very sensitive to their loyalty. There is a lot of details. If you cancel the trip, you need to get the money back, and then the financing needs to reconcile it. Very manual. When we came to the industry, just bringing these things to a browser, was a thing. I remember, and [Grayson] was an early backer, and I remember I talked about it with Mark, and he was like, "You can do AI and you can do machine learning, and all these things." This is in 2017. I said, "No, we are kind of disrupting the phone." We are moving it from the phone, people are calling an agent, to a browser. Now, why is that important? Because once you take the time to automate everything, to basically create an API for everything, I can change a flight. I can get it reconciled. I can get the credits back. I am connected to everything globally. You need to connect to buy a flight in Brazil, that is the acquisition in Brazil. You need to have an IATA license in Brazil. You need to have a point of sale in Brazil. You need to have a United Airlines deal in Brazil. It is very manual, very fragmented, but we wanted to bring it online and also wanted to have it efficient, so you will book your flight really fast. You will get support very fast. Before GenAI became a thing, I am talking in 2017, our first team was a machine learning team, and the idea was to really get to know you, to know what are the flight that you are likely to book, what is the hotel, what you are likely to do in the system, and really show that hotel to you, that flight, that seat. We have built all of this infrastructure. We also had a good AI team. When GenAI came, we said, "Okay, let us go." Now we can take the level of service that we had back then of, I do not know, it takes 10 minutes to book a trip, to now it will take you seconds. It will take you minutes. We will always get you. You can talk with our chatbot and change everything in minutes. AI was always this developing stories from 2016 all the way to now, and you can see us keep innovating there. Ava is a chatbot. It is not like your AT&T chatbot that you call and you get a representative and so on. This is someone that you can have a discussion with and say, "Book my flight, change my flight. I want this. Book me a restaurant. Do all of these things for me." And you will think that you are talking with a person. And in fact, sometimes you are, because we are completely orchestrating the AI agent with the person, and sometimes somebody needs to call the airline. Somebody needs to do something with the receipts. We really perfected AI on the application level to create value for our users and our customers. Great. Last night you guys reported earnings. You announced the acquisition of BoomPop, an AI-native meetings and events company. Just talk to us about the thought process behind this acquisition and how it fits into your broader M&A strategy. Yeah. First of all, I will clarify because I think it's important. You know how it is. We talk about AI, so people go to ChatGPT and asking ChatGPT, so what is the revenue of BoomPop? There was some confusion there between GBV, which is basically the cost of the hotel room, flight ticket, to revenue. They are not the same. Usually, it's 6%-7% out of the GBV. So people got excited that BoomPop was probably doing $100 million. It's GBV. I am clarifying this because then people really got confused with our numbers. But the really important part about this acquisition is actually what it gives us. Meeting and events is a big part of the way that companies are thinking about business travel. You have the transient travel, I am a sole person, I am flying to New York, and then continuing to Paris, but there is the sales kickoff. There is the teams offsite. There is all of these things that people are doing when they are coming to an event like this. Somebody in Goldman, and probably a pretty big team, were working on this event, and they teamed up with this hotel, and they made sure that there is some food there, and all of these things. If you go to the Goldman team and ask them, so what does it mean? They will tell you a lot of phone calls, a lot of emails, a lot of contracts just for these meetings. So very complex, very old school. You book events today in the same way, whether it is a team event for 10 people or an event for 1,000 people, you book it like you are in the 1980s. We, for quite some time, are getting more and more excited about this use case. Navan, we have our own functionality. We call it team travel. This is up to 50 people. When you want to book an event from us today, events like this one, you are actually using Reed & Mackay. The Reed & Mackay team has an events team, and they will partner with you, they will contract with you, and they will do it like any other old-school travel agency. The question is, and the opportunity is, can you automate it? Can you bring it to the AI era? BoomPop developed really cool technology, that you can actually go and do a conversation like in ChatGPT and start to plan the event. I want to build an event in San Francisco. There will be 1,000 people there. I have some ideas of the level of hotels. Here is the event. Go to the other side, to all of the vendors, and start to ask them for prices, for agreements, and so on. They completely automated the two sides of it. We partnered with them, so that our customers like this technology, which is usually a pattern for Navan. We first partner, and then we kind of see that it is working from a team's perspective, but also go-to-market perspective. Now we think that there is a huge attachment play here. We already integrated because we partnered, and now we can basically open our Salesforce and give our sales team the ability to sell all of these events to all of our customers and also excites new customers. You can land and expand with an event and then take it to the other side. So we see it as a huge opportunity to tap to 30% of the top of this travel market. Yeah. Maybe I can add here the great opportunity that this represents for us in terms of growth next year. We believe this acquisition is going to be accretive in fiscal 2028. For fiscal 2027, given the sales cycle and the fact that we just acquired them, we expect maybe a couple million dollars of revenue, and some costs. So, no significant impact. No $100 million. Yeah, no significant impact. We are very excited about this opportunity because the combination of this great technology and our very broad base of enterprise customers, we believe is going to create a lot of revenue synergies. Great. Another thing you disclosed with your results yesterday was that new signed GBV was up 60% year-over-year to $4 billion. Yeah. How can we contextualize that $4 billion number versus your business outlook? Yeah. Very excited to share for the first time, this new metric. It is a very important metric for us. Let me step back about what this is exactly, right? Every time our sales team go and sell to a new customer, usually large enterprise customers, we estimate how much bookings they are going to process on the platform when they achieve their run rate, right? That is what this $4 million is. It is the sum of all the new contracts we signed over the last 12 months, and that is the annual run rate of all those customers. Remember, those customers usually sign for three to five years contract with us, so the $4 million GBV is just one year of run rate of those customers. From there, we have been explaining that those customers are ramping, so when we sign the deal, we pay a commission to our sales team. That is what you will see in our marketing and sales line on our P&L. We pay a commission to our sales team. From then, we help our customer deploy. Usually it takes roughly two months before we start seeing the first booking on the platform. From there, five to six months before we achieve the run rate. That is at this point that the $4 billion start compounding for the next 12 months, right? Because it is obviously a run rate. As you can see, there is a lag between the moment when we sign those contracts and the moment when we see the bookings on our P&L. This is a very key indicator of future growth for us, right? This is the metric that I use to make sure we spend the right amount of marketing and sales. That is the metric that I use to measure our payback and how fast we are going to get a return on our sales and marketing investment. That payback has been very, very attractive, right? That is why you see us leaning into marketing and sales, because of that payback. We thought that sharing this number was an important part of the puzzle when you think about the way we make money at Navan and the way we invest money to generate future growth. That number keeps accelerating, right? That is what we have seen. It is up 60% year-over-year, but just the Q2 portion of that number was the highest ever. We have reported that Q4 ratio was 50%, now it is 60%. We keep seeing more and more customers joining the platform. And actually, not even in that number, we just signed the biggest deal ever for the company. So the sales team keeps really over-performing, and we have great success in the market today. That is what this metric is measuring. Great. I think one of the things that investors are debating today, and we have gotten some inbounds from our clients on this, is the fact that the revenue guidance for the year assumes pretty decent step down in the growth rate off of the front half of the year. So can you help us think through what are some dynamics that could be impacting prior year compares? What is conservatism? Just help us think through the guidance. Yeah, absolutely. I just described this ramp for our customers. So that impacts, obviously, our growth rate, depending on how fast some customers are going to ramp on the platform, depending on how many customers we are going to launch in a given quarter. As we scale this business, that impacts the particular growth rate of a specific quarter. So that is the first part of this answer. Secondly, we just raised our guidance for fiscal 2027 to 32% year-over-year revenue growth, which is obviously already an acceleration versus the 31% We reported last year. So when I look at the last 12 months of bookings and revenue and the trend in the business, the only thing I am seeing is acceleration. So we are very, very pleased. We just discussed about the $4 billion of unsigned GBV that is going to further accelerate the business going forward. So, the way we are thinking about the business trajectory today is clearly acceleration. And then to your question about conservatism, maybe, that is not the way I would qualify our guidance. We like to guide to a number. We have a high confidence we can achieve. And, as a reminder, we are a usage-based business. So there is a lot of things we control, some things we do not control. We make money when our customers use our platform, and that is why it is so important for us to deliver the best service. We will see. It's still early in the year, there's still five months to go before the end of this fiscal year. But we are very, very confident, and we really like what we are seeing right now in the business. I will add to this, without getting to the mathematics of, obviously, 32% is bigger than 31%. I think that I'm always asking myself several questions. How do we do in SLG? And we just described, if we are doing well, it gives you a good visibility for the next 6- 18 months because of the run rate principle there and so on. That's the first question. The second question, how are we doing with PLG? We didn't report how much money this adds. We reported the $4 billion GBV. And PLG has much shorter cycle, so therefore it translates to revenue really, really fast. If you think about the size of Navan of net new, it's actually bigger than $4 billion because of PLG. So I'm asking, is PLG doing well? And I said it's actually more than doubled. By the way, it's always more than doubling, so it's actually growing really nicely. Then how are we doing with payments and expense? When we went public, it was a business that was decelerating, and actually now it's accelerating. So we're doing well there. And how are we doing with our new bets? Navan Edge, Navan Anywhere, new products. And from where I see it, again, without getting to the model and the numbers and all of this, all of these things are doing better than what we could have imagined a year ago when we went public. So I think this kind of gives you the hint of how I'm seeing it. Now, it's also important, we are a new company in the public market. I think it is important to be very, very prudent in not making a mistake in what we are reporting and kind of be responsible there, and that's why you see 32% guide. Okay. You mentioned some of the new bets. If we think about your AI roadmap, what has accelerated in your R&D approach or your conversations over the last 12 months? What are you hearing from customers that they want from you, and what are you betting behind? I think the one thing that surprises me, how fast it accelerates with our team and how much it impacts the quality of our discussions, I am talking about both discussions, is actually the use of our model. It was interesting. It was even a debate like eight, nine months ago, if we really want to go to our model or use more frontier models and also kind of open source. I didn't think, and also Ilan, my Co-Founder, who is the CTO, we didn't think it will be so impactful. 50% usage of our own model with Ava impacts first thing actually CSAT. The satisfaction of our customers from the discussion. The discussion is way more relevant. It's also a faster discussion because I don't need to go to somewhere else's and to be on a line, kind of a virtual line with everybody else. This is the reasoning that you see all the time. Right. I don't need to do that. Right. It is faster. It is much more relevant. Obviously, cheaper. This is when we reported now we are at 50%, and it is growing. This is for us a huge investment, but also very surprising, the outcome of this thing, and it impacts across the board, right? On every AI product in the company. The next thing, Navan Edge. While for me, it is obvious that people will want this kind of EA or travel agent in their pocket that really knows that they like the Marriott, and they like this seat in the United Airlines, and they would probably like this restaurant and this, I do not know, the Lady Gaga show in Madison Square Garden. By the way, all real use cases. For me, it is kind of obvious, but I am techie. I like to think that people will actually do that and see them doing that, and see them do that in a product that we have never had a product that accelerates so fast in terms of usage, and actually more importantly, repeating usage. I booked once, and I am coming in the next month to book again. I think the potential of Navan Edge, we are just getting started there. Now, it is early. We are not sharing numbers yet. There is- Right. a lot of engineering work there, a lot of product work there, and probably most of the time that I am spending on products is around Navan Edge. But it is really, really exciting. That is exciting to hear. It gets into the fact that the bar for consumer experiences is going up with AI, and it feels like that is translating to an extent to the enterprise world as well. I know historically ripping and replacing incumbents has been difficult, but maybe the pain points are getting bigger. Maybe there is a modernization push. Is this something that, as you guys go to market, as Michael talks to customers, is this something that you guys are seeing? 100%. If you think about it, I'm also running a company, right? I have an IT team, and I'm coming to the IT team and asking them, "Look at this kind of SaaS product." I'll not get to names, and are you sure that that's the product that we want to continue with when they renew, right, in the coming years, given the technology advantage that I can get with AI? So I as a CEO is actually coming to our CIO and telling him, "Hey, you need to, before we renew this," and I'm talking about products that people have been using forever, right? So like CRM, ERP, HR, and I'm asking, is that the right product for us going forward? If it is, great, we are renewing. If it's not, we are looking at what's out there, and maybe out there is startups that are still not ready for us, but we are looking, right? Navan, in this case, is very unique. We are very scaled. We can support companies like Unilever and Netflix. So you know that we can support you, but we are very AI. You can connect Navan today through MCP to ChatGPT and ask, what was my spend six months ago on travel, and who is the employee that spent the most, and how can I save, and whatever you want to ask ChatGPT. This is super powerful functionality for a CFO, for a finance team, for a manager, and I'm not familiar with other kind of companies in this space at our scale, at our capability to support an enterprise that can do that. So again, it gets back to those enterprise customers want AI, hell yes. That's exciting to hear. Yeah. Give us your thoughts, and I think you kind of alluded to it with MCP connectors. Give us your latest thoughts on headless. Headless was always before GenAI, right? It was always kind of this discussion in Navan, should we open the platform for everybody who want to build some startup around business travel, connect to us, you want your travel agency connect to us, or you want to appear in a lot of places. It was always a discussion. Also, some customers that really care about the experience of their employees, came to us throughout the years and said, "I want Navan, but I want it with my logo, and I want it kind of with my colors, and I actually want it in my portal," right? It was always a thing. The problem is that before GenAI, you could open an API, but in travel, you have a lot of skills. You really need to know it's not booking a United Airlines flight. It's in this class, and can I change it or not? I'm talking the third class, and these are the rules, and all of this knowledge that is there, and GenAI allows you to pass it, right? You obviously pass the ability to book a United Airlines flight, but with all of the knowledge that a travel agent has. So when you connect to us today, you get all of the goodness of this knowledge. But at the end of the day, when you want to change a flight, you can still come to Ava, change the flight. It's all connected, and now you are basically dealing sometimes with a human agent. So in a way, AI allows us to pass this functionality to where people want to be. It could be email, it could be Slack, it could be Gemini, it could be whatever, and enjoy it while you're not giving up all of the things that usually tend to get very complex when it comes to business travel. I'm very excited about it because it just meets people where they are. They can use the full-blown Navan Edge or go and just consume it from Gemini. Do you think, and I know it's very early days, but how do you think it does it expand the TAM? The TAM in business travel is so big, and we expanded it anyway because we have this slide that we use during the roadshow in the IPO, which basically put the user in the center. Then you kind of have Cognition, Navan Cognition is AI platform, and then there are all of the needs. The needs could be transient travel. I'm traveling to New York. The need could be I need to pay for it. I need to expense it. We've mentioned there, and we said VIP one day will be on platform, and that's the Reed & Mackay move of let's move VIP to one platform. Now you can talk with a VIP agent, but you can also book it online. You can change it with Ava. We just announced meeting and events, right? The TAM is so big. Each of these things that I've just mentioned is tens of billions of dollars. It's so big that I don't know if AI makes the TAM bigger. It makes the advantage of Navan over the competitors almost not a fair advantage. I know who are we competing with. We are basically competing with, I'm calling to an agent, or I'm sending an email, and we are basically here. I think it gives us a huge advantage of the ability to take the entire market. Does it make the TAM bigger? It's such a big TAM. I don't know. Fair. One of the things Navan often cites is its partnerships with travel suppliers as a key point of differentiation. I know you've done a lot of work hammering out all these agreements. Can you talk to us about how this works? What advantages do these relationships give you? Yeah. From day one, you know that you use more of a legacy B2B system for travel. There is this ongoing complaint. I'm going to the system, or talking with a travel agent, and the travel agent is telling me, "This is the flight, this is the hotel." People in business travel are behaving like consumers, right? You are traveling. You care about which hotel you're going to stay in. You care about the seat. You care about the airline, right? You are actually like a consumer, although the CFO told you have to use that. So they're getting really upset, right? They're basically saying, "You're telling me to use the system, and the prices in the United Airlines website are cheaper," or, "I can find a cheaper hotel online or in Booking.com or whatever." That was when we went to the industry in 2015. That was the common complaint. You are basically telling me to use an inferior system without the right prices, the right content, and so on. From day one, we said, if it's out there, it will be in the system. This flight, we are the only ones that are supporting low-cost carriers, like easyJet. Why are we supporting it in business travel? Because people actually use it for business travel. So, low-cost carriers in India, stuff like that. We have this principle, if it's out there, we are connecting to it, but also we are providing all of the information. Information could be how the seat look like. Delta One versus Polaris versus this thing versus that thing. There is a picture, though. This is how the seat look like, and the angle will be this or that, and people really like this seat. You cannot get this information from one system. You need to combine systems. Sometimes you need to connect directly to the airlines or to the hotels. Sometimes you need to bring the content from a GDS, Sabre or Amadeus, but they bring the prices from NDC, from going directly to D. Our users don't see that. But for us, we are connected to everything. We take all of the data, we feed our model with that, and then you can come and say, "What is the best seat for me to sit in a United Airlines flight?" Or, "When I fly to London, what's the difference between this flight to London or that flight?" That's why NDC is so important for us, but GDS is as important for us. We are connected to everything. I would add that, conversations with our partners and suppliers, we are bringing a very important demographic to their business. Business traveler is a very appealing category of traveler, obviously. As we scale this business, we are more and more important. I know many of you are coming from New York. In every flight between here, San Francisco, and New York, there is on average six people who booked on Navan. We are getting bigger and bigger, and we are bringing a lot of important business to those suppliers. Yeah. On that point, you discussed GDS versus NDC, both are important. What are the economic implications of both of those booking channels, and does offering NDC become more commoditized over time? The technology is evolving fast, and more and more companies are adopting it, but it is still early days, and we are by far the most advanced in the space. We just released the first NDC connection with a hotel chain, with Hilton, recently. That was the first in the entire industry, and we are by far the travel agent that has the most connections. It is just the way we have been building the business, as Ariel mentioned earlier. On your questions around the economics, there is not an easy summarized answer to this, because every arrangement we have with every supplier and every GDS is different. Obviously, direct connection, the biggest benefit is we offer a live inventory and broader inventory and live pricing, which I think makes it way better for our traveler and our customers because we offer the best prices. That is one of the reasons why we are able to save them on average 15%. Not the only reason, but one of the big reason why. I think it is very critical. It is a marketplace. In this marketplace, you have the suppliers, airlines, hotels, and so on. You have the employee that travel, and you have the company. We always thought that there is a win-win-win between the three, and everybody here needs to get value. If they are getting value, we are going to get paid. If the airline is getting value from this marketplace, the airline will pay us, same as the hotel. Actually, the Marriott CEO yesterday, in a completely different context, was talking in a conference, and he was saying, "We are giving some preferred terms to websites that you can do a conversation with because we think that that is where the market is going." What I have learned over the years with airlines and hotels, they do appreciate technology when it comes to experience and also when it comes to representing them correctly. They really do not like it when you are saying, "This is all business class." They really care about. Somebody spent a lot of time and money to think about Delta One. Somebody spent a lot of time and money to think about Polaris, and I can go on and on. Same goes with Marriott, same goes with Hilton. So they care about it. We are really good on representing their product in the best way possible, and to create the match with our users, our travelers, and with the company. This is why eventually we get paid by the various kind of people in the system. Yeah. Maybe following up on that point on differentiation. Amex GBT is trying to execute an AI-enabled turnaround. We get a lot of questions on this. What are the limiting factors to competition catching up to you? I think, first of all, when I am reading the press releases with Amex GBT and others, first of all, it is fun. It is the vision that we have been talking about for the last 11.5 years, and at least on a press release level, everybody are now doing it. It is actually fun to see. I think, traditionally, our competitors are employing a lot of travel agents to eventually book your trip for you and then change it. It is rare that we will have a travel agent that will actually book your trip. We will have travel agents that will change your trip and so on, and even then, it is with this orchestration that I always talk about between AI agents and travel agents. To take this knowledge that is basically in people's brains globally in every other language, and then to create a model out of it, I think it is a really tough engineering problem. I think they have the vision to do it. They will need to create a lot of things. First of all, an online infrastructure. Our competitors are very global. They have licenses everywhere. They are operating everywhere. But this is physical office with travel agents sitting inside, picking up the phone or answering an email. We have created an infrastructure that is basically connected and doing it automatically, booking your trip, changing your trip, all of these things. I think that it is very complex to build this infrastructure and then to create a model around it, and then really understanding the UI for companies that are by nature not tech companies, right? First of all, I am excited because they are definitely now finally speaking our language. Execution, I think, we have competition, and we will make sure to stay ahead. Great. We have a few minutes left. I would like to open it to the audience if anyone has any questions. Otherwise, I am happy to keep going. You are very excited about TravelClaw and Navan Edge, as are we. I guess as we think about agentic workflows that offer some compelling demos, we saw an example at your customer conference. Looking forward, what are the most important KPIs investors should watch to determine whether this idea of autonomous travel management can become mainstream? Yeah. I think the one that we started to report is actually model usage and what do you get out of this model, right? It's not just the model. When we are saying 60% of the interactions with Ava were done automatically, it basically tells you, and this number kind of keeps going up, right? If you compare it to the path, it basically tells you the value that you can get from AI on an application level, right? I think watching these two numbers will be very important, and kind of watching, and it will take some time, what I'm saying about Navan Edge is also very important because this is really the application layer. So we get value, big economical value today at scale today from Ava. You see it in our gross margin. It's rare to see, travel is service, right? It's basically service. To see a company that by definition the business is service doing 75% gross margins, it's rare, and it's pure AI. There is nothing else that creates it but the infrastructure that we basically API'd everything, and then connected it to GenAI so you can have a discussion. These are the two numbers that if I'm an investor, I will watch carefully. How much of our model we are training, how much we are basically having AI discussion, 60%, therefore, these are the gross margins. Great. I think that's all we have time for. So please join me in thanking the Navan team for their time today. Thank you. Thank you so much. Thank you. Thank you