Fireflies.ai's path to $1B valuation: How being "too early" to AI paid off

0:00Probably right now, the biggest gains are happening in coding and customer support. that's where these AI companies are taking off, getting insane valuations. The desire to make coding more accessible to everyone. I would love to see that same sort of AI play out in AI like collaboration, workplace collaboration stuff. So that's what I'm most excited about because there hasn't been that breakout company yet. And we would like to be one of those. Welcome to the SaaS Revolution Show, a podcast by SaaS.com. Here, we interview SaaS founders from around the world who've been there and done that, as they share the ins and outs of how they built their businesses, their operations, their path to securing investment, and more.

0:44Our mission with the podcast is to help you, the founder, learn how to scale your SaaS, maintain your wellbeing, and navigate the complexities of this ever-changing industry. I'm your host, Alex Deema, and together, we'll explore the good, the bad, and the ugly in the journey to SaaS success. AI-driven, ARR-focused, the SaaS Stock Europe 2025 lineup means business. Hear from the founders and leaders defining our industry, including Manny Medina, co-founder and CEO at Paid, Nicola Merksic, co-founder and CEO at Poly AI, Alice Carlisle, regional VP UK and Ireland at Wiz, and Stan Masueras, GTM Europe at 11 Labs, ready to build alongside the best.

1:35Book your tickets now at sasstock-europe.com slash tickets. Welcome to the SaaS Revolution Show. Last week, I sat down in London with Krish Ramaneni, co-founder and CEO of Fireflies.ai. we got into how he built an ai native sass company before the ai boom why fireflies is moving beyond note-taking to ai teammates and what the future of sass looks like in the middle of this ai bubble let's dive in so yeah so i definitely think like london and sweden i'm seeing those those as kind of like pockets and where you know like all these big rounds are just constantly being announced on a daily basis um you know all ai native companies of course um but yeah very interesting time so i guess most of the uh ai native companies they're like they've been created in the in the last year right um and i guess kind of leads to like probably the the first question because when we we look at open ai as we think starting the ai you know revolution and i think that's quite fair uh the interesting thing and you know doing a little bit of like homework on you guys is that you know fireflies has been around for nine years uh i've listened to a podcast you did in june 2022 and open ai gbt1 you know came out in uh whatever november 2022 i i think it was so i was sort of curious to understand how you're so like prescient to like create an ai native company nine years ago uh where obviously it's much more like commonplace now and how that uh helped or hindered you because it it's only now we're seeing the boom of ai and how commonplace it is yeah well it's been quite a journey right 2016 was when we started working on a whole host of things.

3:39And I always like to say that we were a couple years too early to the market. It's a lot of trial and error. The other big part of that journey was there were many products we launched at that time that were too early to market. Technology wasn't good enough, but now would be perfect. So we will be going back to those and building those into the larger Fireflies suite. Today, people know Fireflies is the AI meeting assistant, joins your meetings, takes notes for you. But before the meeting assistant, there was an email assistant, there was a Slack assistant, there was an AI task manager. All those things was built in the pre-LLM era.

4:17Today, as we've worked on meetings and scaled that up, our thesis has always been the same, that there is all this knowledge buried inside conversations. So we want to actually bring those things back and build for all of those different conversational streams, not just meetings. Like where are the other places you spend so much of your time outside of meetings. It's usually emails and Slack. So those will be the first areas that we're planning to tackle. Someone said this as I was talking about this longer term vision for our company is, hey, you're essentially building this like cursor for conversations where you're capturing all these conversations, unlocking value out of it and like executing on those conversations.

5:00So but yeah, the first couple of years were very hard times. In fact, I'm curious, when was SaaS stock first started? 2016. So we were kind of, well, if you look at, I guess, SaaS dating back to year 2000, right? But it did feel like we were, or that era, 2016, it was still quite early in SaaS, even though it was 16 years old, right? And obviously, we're early in AI, right? So I wonder if that will play out similarly. But yeah, like the first SaaS stock, we had like 700 people there flying in from all over the world. And you look at it today, it's, you know, three and a half thousand in our European event and our US will grow similarly.

5:47But certainly it's kind of ridden the coattails of, you know, the SaaS wave. And then you look at the last 23, 24, there were headwinds in SaaS, right? Yeah. And then we naturally sort of feel that. But AI, you know, which, you know, considering it is like, you know, certainly the application layer, B2B software is bringing back this great tailwind, which I'm also very thankful for. Well, I have a really fun connection. When I first started Fireflies, I didn't know what SaaS was. So I was literally going on YouTube, typing in SaaS. And one of the first videos I saw was SaaS stock. and I would keep watching those SaaS videos like throughout 2016, 2017, 2018 because we were trying so many different things and it wasn't working.

6:37And actually some of those videos from the conference were super helpful because when we first started we were going upmarket enterprise and then this is not working. Like people are not adopting this stuff. And then I remember vividly seeing a speaker talk at sas talk about selling the smbs how they made it work and they broke all these like conventional rules which you typically have as you need to have sales people you need to do this stuff and that actually set in motion a lot of how we build our company today so also it's a byproduct of i'm trying to think i mean like there's almost there's like an accounting software company that was very prosumer there was like cal and libel a bunch of these other uh founders that came yeah and talked about their journey.

7:24It was very different from all this other startup content that was out. All the literature in 2016 and 2017 was very much sales-led because all the people that were writing blogs and content were sales tech companies and marketing ops companies. And so you thought that's what you needed to do. You needed to set up a big inbound ABM, all these like sort of funnels, whereas PLG wasn't being written about a lot yet you had great companies at the time already that have established themselves like DocuSign and Calendly and Dropbox all of those sort of companies and I felt like in Silicon Valley it's a very much an echo chamber where in 2016 2017 2018 every person we talked to every advisor every guru out there would be like why aren't you selling top-down enterprise sales that's how you make money you got to get high mid-market or top-down enterprise sales and i said we're building something in a brand new category this category doesn't exist it may not even be a painkiller right now it's a vitamin and we need to go to people that are going to give us valid feedback really quickly so my gut was always telling me smbs is the place to start and seeing some of your guys's content actually helped well that's uh that's great to hear and i think like um i i guess kind of uh topical but i remember and actually i was just what's happening somebody who's looking at coming to sas stock europe which is taking place uh next month in october he was asking me for some stats on the the attendees and we saw that um with our new data platform that we can actually really capture a lot like understanding who the attendees are and 55 percent of companies attending sas stock are plg and i imagine this is you know uh it's become the trend over the last two years for sure but like one year ago i didn't have this data and i was being asked by uh you know a long-standing sponsor who had this perception that everybody was enterprise sales led and i knew in my gut that actually i'm speaking to so many plg founders there but i just can't share the data point because i didn't capture that information so it's great it's great to see like i guess kind of on that topic like i remember it's got to be like a couple of years ago it could have been 2022 when you first like maybe it was one of my sales guys or i was on a sales call and fireflies like pops up and you're like what is this doing here is somebody spying on me right um oh i don't like this right because it was very new and quite often booting like the firefly like do we need this right and booting that out and I imagine because you know that was a new experience it's human behavior there was probably you know some of these challenges that you were seeing or the feedback that you were seeing and then so I'm curious to hear about that and obviously as you fast forward to today you know I was on a call yesterday with a colleague and you know we've got note takers internally and I could see she's writing in her notebook why are you doing this right yeah you know you just put the note taker on um so it's interesting to see how in certainly in the space of a couple of years how certainly our behavior and human behavior has changed like how have you seen that and shaped that and how has that helped you with your growth when i look at my own behavior i realize how dependent i am on ai every single day whether it's chat gpt whether it's fireflies or a dozen other tools.

10:59I actually make it a point to try at least two to three new AI tools every week. And maybe not directly in our category, but other categories. I'm absolutely amazed by some of these video and image generation tools like Higgs Field and Nano Banana. It's just so many great tools out there. HeyGen is another cool company. Eleven Labs. We like to figure out how to scale our marketing. So I'm trying out all these tools. So there is definitely a proliferation of tools. but the hardest part is there is only a finite amount of attention that is out there and people are constantly overwhelmed with so much app bloat and if you look at all the apps on your phone too there's probably only like three or four that you use every day so we were fortunate enough to start at a time where this was very novel today in two to three weeks anyone can build an AI note taker.

11:54Like it's not that special. Back in 2016, 2017, 2018, we had to solve a hard technological problem, transcription, summarization. We were working on like neural nets, all of those sort of stuff, along with NLP classifiers. We had to do everything from scratch. Now it's just an open AI call for a lot of the things that you need to build. And transcription has relatively been a solved problem today. And now we're at the edge, right? Like if it's a noisy environment, a windy environment? How do you improve on that? So there's so much more things that you can take off the shelf and build. In fact, when we were raising money, people thought like you couldn't build a company with off the shelf stuff, you need to build these like proprietary models, you need to hire these machine learning people from Apple.

12:38And this is how you build a company. Obviously, as a small startup, that was not going to be possible. Going back to your first question on even before, like we had to solve all the technical problems. We launched in January 2020, A few weeks later, COVID happens. Everyone's hopping on video calls, Zoom and Teams and Google Meet meetings. And all of a sudden it went from this vitamin to a painkiller because you're having Zoom exhaustion. Like you don't remember what's happening on those meetings. And I always like to say that our marketing and our pitch was really about today, the transcription is not that great, but tomorrow it will be amazing.

13:16This is what we said in 2020. And today, when we look at the customers and the feedback they leave, less than 5 % of them complain about transcription quality. When we first started, the most common reason people would churn was, yeah, it's not accurate enough. Like, I don't think I can rely on this. So that's really betting on the tech. The other part was before ChatGPT came out and like the LLM APIs came out, I think around GPT 3, 3.5, that's when we started using LLMs as a way to start summarizing the nodes. and we were amazed at how much we could stylize it, how much we could personalize it.

13:54And we had this dream in 2018. I remember reading a research paper that came out of Salesforce about neural nets and how that might be the new move instead of just using regular language processing. And I remember telling my co-founder and CTO, Sam, maybe one day AI will be able to write better than me and AI will be able to paraphrase and create notes better than me. Because us as humans, we get distracted on meetings. We don't remember everything. Our recall is really bad. So after a meeting, you probably forget automatically 50 % of what was discussed. And then a couple hours later, after a few other meetings, you probably retain only 10 % to 20%.

14:33So let's solve for that problem. And when people were starting to use Fireflies initially, it was like, let's give you perfect memory. It might not be perfect transcripts, but you will have a really good enough notes, transcripts, and a search component where you can search back through all of your conversations and remember what you talked about. And the use cases for that are infinite, like what you can do that will help you recruit better, they'll help you manage your relationships better. So all of those sort of things was very novel at the time. And we also had a different mechanism with which we were trying to build this company.

15:07And today they call it agents, they call these like browser bots. But back in 2017 2018 we were already building this tech stack and we wanted it to be so seamless that you don't have to like go hop through 50 different steps to set this up and the ai note taker bot we really pioneered like a bot physically joining your meetings using rpa like robotic process automation and i always like to tell people that while fireflies is an ai company there's a whole lot of robotic process automation that we've had to build like all this browser tech that we had to build to do what we do today. In fact, there's now like an API that other companies provide just for you to build bots, right?

15:48So everything can be done off the shelf. It's funny how what investors were concerned about almost seven, eight years ago are things that you can now pull off the shelf, right? But on the social stigma part, it was definitely very challenging in the beginning, because you have to like explain to people, oh, why are you using this note taker? Like, what is it doing are you trying to catch me to say something but today it has become more socially acceptable and the one nice thing with fireflies is at least it starts a conversation and you see the bot there it's not like someone's secretly recording you in the background like some of these other tools do uh which a lot of enterprises and a lot of larger companies don't like uh the other part is that compliance piece where hey i'm using this to take notes and record my meetings if you're if you'd like i'll send you the notes afterwards and we had people that were strong critics turn into wow this is actually valuable i want to start using it for myself i think we've changed thousands of people's perception over the years from i would never use something like this to i can't live without it you mentioned the the proliferation and there is this saturation now and it probably will be even greater of all these ai tools and you know lots of different note takers out there uh being first to market probably you know almost certainly has given you you know some advantage but how do you think about it from the defensibility you know from the moat sort of like perspective when you know i don't know how many note takers there are out there now but there's probably hundreds now yeah yeah yeah well we've had a long time to work on this problem and continue to iterate on it, like from the summaries to the transcripts, to the user experience, to administrative features.

17:35There's a whole lot that you can do to go really, really deep on these problems. In fact, we had a customer the other day that came to us and said, hey, there's this other company. They're like an AI CRM company, but they really started off as an AI note taker. It's almost like the note taker is the entry point for you to build almost any other SaaS products, if you want to build a great CRM today, you probably need to have an AI note taker or integrate with Fireflies. So you have all of that conversational data. I feel like we add more value to CRMs than any other data source out there, because we have all of this rich contextual knowledge about what's happening in the meetings with prospects and customers that actually make the CRM useful.

18:17Otherwise, CRM is just going to be rows and columns, right? So when we look at all of this stuff, we go a lot deeper on integrations, we have 70 plus different integrations, we were the first meeting assistant to connect to Cloud MCP, as well as ChatGPT connectors, as well as Devon. So that accessibility, we're probably one of the only notetakers with a very robust API. And that allows people to use Fireflies in a lot of different settings. We've also done so much work around teams, because there's a lot of prosumer individual oriented notetakers. But for teams, you need different things, especially sharing access controls, data retention, user groups, all of those sort of configuration stuff.

18:58Well, recently, there was a note taker company that got in a lot of trouble. And there was a lawsuit that's ongoing. I don't know the full details of it. But the general consensus was, this note taker company is using our data to train our models, which is something that would make a lot of people squirm. And I'm sure many other companies are doing that. Whereas our credit to our security team and our CTO for coming out proactively several years ago and saying, we're going to take a no training on customer data policy first and foremost. So winning trusts, winning the right type of customers, integrating deeply and building for teams.

19:36So today, Fireflies doesn't train on your data by default. We have a private storage offering for larger enterprises or even for small companies that want to store their data on their own servers. You own your data, zero data retention policies on like LLM models training on your data. We don't have that. So we've had to do so many different things. And yeah, I don't think like, you can do that in a matter of weeks where, okay, I'm just going to build like the note taker, you have to do a lot of these other things. But we don't want to stop at the note taker at all. Like the bigger vision for Fireflies is that we want to help you unlock all of the knowledge inside your conversations meetings is just one data source.

20:16We want to bring that value to your emails, your Slack, and we want to bring that value to all the work that you do. And till now, we've just automated meeting notes. So I had this idea that what if we automated the meeting itself? So we're also building AI teammates that can go take meetings on your behalf. It's building like a full-fledged platform. I'm just glad we picked one use case with the meeting note taker which ended up being the defining gen ai use case like if anyone wants to build a company today in the sas space uh one of the top three things that'll probably come up is building a meeting note taker in fact there was like a company that made a lot of noise on social media with their posts and controversies and skits and stuff uh but eventually they all pivoted and ended up coming down to oh you're just another glorified yeah i know taker so it's uh it's funny to see no it's interesting um and a couple of things you said there because as you described it and you know i'm thinking of like fireflies as like a second brain right and in in some aspects of the ai tools that i've played with right you know with 11 labs i've been able to clone my voice so i don't have to do podcast adverts necessarily and it helps save me some time um uh we've got a um we're using a tool called real flow on our website uh where it's like videos of me you know kind of explaining you know the different products across our pages but i think through using hey gen uh we've cloned uh myself so that if we need to do any future edits and updates that i don't have to sit there and do the recording which i the initial recording was like three hours right of like reading scripts and stuff like that so again it's saving me time for future um cloning me so i don't have to do meetings i mean i love that idea you know if that's possible because i think a lot of people don't love meetings you know some meetings are like a sales meeting is it's very useful and you know if somebody says oh you've got a sales meeting tomorrow you know i'll i'll take that a lot of the internal stuff um you know sometimes it can be overdone and it's not the best use of your time but finding a good way to clone you know a person uh to do meetings definitely very interesting um so i'd like i'd like to see that there's at least 20 of meetings that shouldn't even exist yeah and uh some of them can be automated away especially like screening interviews the initial screening call if you're doing the same call over and over again certain types of customer support calls that can be automated even like sales, a rep demo discovery calls, a lot of these Q &A type things where you're gathering information, user interviews, screening interviews, these sort of things.

23:07I think we can literally automate out that first initial meeting. If you do that, think about how much time you would save for someone. Also, I believe that if I have to wait X amount of time to go, let's say I want to go try out this like software, but I have to book a demo with them. And it will take me like two weeks to get on a call. I got to fill out all these forms and stuff. Whereas if I have like an AI sales rep that can give me the demo and I can ask questions right there, it'll do like 90 % of the thing for me. Like I can make a decision a lot faster. If I have 50 candidates that are applying to my open role, like we have 500 candidates applying to an open role today.

23:47That's a lot. And it's going to take our recruiter to go through all those candidates weeks. Whereas we now have our AI recruiter that has interviewed hundreds of candidates in a couple hours and then run scorecards for all of them, stack rank them, given some insights. Whereas like our recruiting team can now go through all of those screening calls and interviews and all that stuff and see who would be the right fit for us. Ultimately, the human recruiter is still making the ultimate decision. But like the information gathering part is done a lot faster through this on the ai sales demo um you probably wouldn't be surprised that this technology exists and we recently through my my fun back future did a small investment in a company called supersonic uh where it does exactly that so it creates the ai sales demos and it's quite a wow moment you know sort of seeing that but just to bring it into perspective most companies are not using you know such tools yet um and i won't name the name but we had uh um i was looking at employer of record technology and going to the usual kind of suspects and uh one of them it took like two weeks uh to get a call and during that time the call kept getting cancelled by the account executive and sort of moved i was like do you not want the business it was very frustrating but you know having to wait that time but eventually i went with somebody else because you know in part because of that experience but yeah as you say imagine the power of just going on the website and just getting the demo there and then getting that information having the conversation and signing up you know if you can respond if you can respond to a prospect within five minutes the likelihood of them going with you and closing them increases by like 50 percent yeah just because of how quickly you're able to respond to them you're catching people right when they need something if you wait too long they might be distracted it just takes them one click to go to another website to another competitor to find what they need.

25:46So we're constantly obsessing over how do we deliver value? And being a PLG-oriented company, we really built that freemium trial experience so that you can try Fireflies and purchase Fireflies without ever talking to us. In fact, we have customers who pay tens of thousands of dollars, maybe even six figures, where they've never actually needed to have a sales call with us, right? I think this year you guys reached unicorn status evaluation, right? Congrats on that. Imagine the virality of the product has driven a large part of the growth. You're on a call, you see fireflies, people wonder what that is.

26:28They start to use it themselves. 80 % of all of our growth came from word of mouth and the viral loop in the first couple of years. Just this year, we started to do some paid marketing. But till now, you could say that the billion dollar valuation came fully on the backs of organic and viral loops. The 20%, then where, like, can you shed some light into what was done there to make up that 20 % and to help with the growth? Yeah, the 20%, at least today, where we're investing is we believe there is value in influencer marketing. uh pr is helpful but today podcasts being able to go to the right influencers founder-led marketing is something i've been doing and we've been scaling that up uh over the last couple years like just like on instagram i'm able to with a post reach tens of thousands if not hundreds of thousands of people uh very very quickly linkedin marketing those sort of things the channels uh are that people thought were like, oh, these are just for fun.

27:34Like if you think about it, like Duolingo got like a billion views on TikTok with their bird. Like marketing is changing and B2B needs to evolve for that. You can't be doing the same old boring case studies or put your email in and download this ebook. Sure, it'll work if your child's selling to top down enterprise buyers. But everyone is like evolving and your buyer may not just be on LinkedIn or maybe not just be on some listserv. they might be doing something on Facebook or Instagram or doing something socially and then you show up. So I think brand marketing is something that we've been thinking a lot about.

28:10Also at the company at our scale, I used to think like, why would anyone waste money on brand marketing? Okay, performance marketing, I get. You spend X, you make Y, you make the LTV CAC and the ROI all make sense. But attention is really, really important. And people, because they see fireflies so much, whether it's in a meeting, whether it's in a video on Instagram or like a reel that I do or a podcast clip, it reinforces that, hey, these guys are legit. They are a big company. And when we did the fundraising or not the fundraise, but like the unicorn valuation with the secondary announcement, I think it led to a lot of people saying, I knew you guys were big.

28:52I didn't know you were that big. So all of those sort of like social signaling and brand marketing is extremely important. that makes up to 20 % of what we do. How much time do you personally spend on like founder-led marketing? I've been trying to do more of it today. I'm glad that we're building a team around it. Being able to do podcasts and stuff, this didn't come organically to me. Like in the beginning, I wasn't like a natural at it, but I think I've been able to be very candid about my thoughts. I'm very unfiltered and I can share whatever I need to. And that's worked really well. Like being on podcasts, doing things in person as challenging as it is.

29:29I think those have been really, really helpful. The other stuff is I wish I could literally, if I had nothing better to do, I would probably be able to just sit and tweet 50 times a day and then post like some inspirational posts on LinkedIn. But you have to have a business to run and you have a lot of these other things to do. But I do feel that I'm trying to allocate at least 10 to 15 percent of my time to this. My team would like me to spend 30 percent of my time on marketing. uh so we're starting to figure that out we're also hiring for a lot of marketing roles today product marketing we're calling this uh brand marketing role vibe marketing but someone that i think in gen z can connect to like figure out how to tell our story in a creative engaging way uh i love some of the stuff that these influencers put out where they're like doing these skits and memes and corporate style skits uh which is really like great i think we can tell b2b brand stories just like how consumer brands do it.

30:28And we can do it in a very cinematic, fun, engaging way. That's an experiment we're trying to do. We actually just shot a couple commercials yesterday. And so I'm looking forward to seeing how those turn out. Switching gears a little bit, do you think we're in an AI bubble? Do you think that's going to burst? And what do you think the impact of AI is onto the future of SaaS? You don't even have to ask if it's a bubble. It's clearly a bubble when you're seeing companies raise at insane valuations. Like the seed rounds today are valued more than like many of the series A, series B rounds back in 2016 to 2018.

31:07Granted, there are some companies that are in hyper growth mode, like Lovable, Cursor, Devin. All of these companies are now in double digit billion dollar valuations. It makes you look at like, what am I doing? and I'm so proud of everything we've accomplished and how hard we worked to get to our first million dollars of revenue and then 10 digits of revenue and then now scaling towards nine figures and beyond. Like that's, but it's been a long journey. It's been a grind and we've broken a lot of the SaaS like benchmarks, but some of these companies are breaking those benchmarks 5X and you kind of look at yourself and it's like, am I doing something right?

31:46We're in a very good position. We're profitable, we're scaling, but you see some of these companies out of the gate in six months to a year get to 200 million in ARR. Like it's insane. And one, there's probably a lot of appetite to the pricing models are changing, like traditional SaaS companies that are doing seat based pricing. I think these all these companies that are doing usage based outcome based pricing are able to extract a lot more revenue up front. And there is a propensity for people to buy that. And many of these companies are not only just selling to enterprises like you have lovable where it's hobbyists prosumers people that want to start like a new uh business and turn uh something into an app really quickly where they're paying you as much as what like a hundred seat or 200 seat license would pay you just one individual one prosumer so the unit economics are also changing i think all of that is good it creates a lot more opportunity it make it democratizes ai ultimately that's what we want, right?

32:44We want everyone to be using AI. It won't be like a place where are you an AI enabled company? It's like asking someone, do you put your business on the internet? You have to, whether you're a mom and pop store or a restaurant, everyone's on the internet, everyone's going to be using AI, it's gonna be plugged into every aspect of your life, your customers are going to be interacting with AI. If you have a restaurant, and you have people calling to make reservations or orders, AI is going to be able to handle those. So AI is going to permeate everything. And while we're in a bubble, while 90 % of these companies may not make it, and while this bubble may pop, you're still going to have some incredible companies that come out of it.

33:23And the net net, the consumer is the one that's going to benefit. During the dot com bubble, absolutely valuations were crazy and stuff. But you ended up having generational companies like Google and Amazon. And then later after the dot com bubble, even with like the 2008 recession, you have generational companies like Airbnb come out of it. So you'll always have like a few companies that make it out, it'll change the industry forever. Like from 2008 to now, the idea of I can just press something, order something, have it delivered to me, go take an Uber anywhere, be able to go to any country and then find a nice place to stay in.

33:57All of those sort of things like happened in these bubble like eras where insane funding was needed, right? I work in the B2B AI work area, right? Like productivity. So I think we're going to see some of the same sort of gains that we saw in probably right now, the biggest gains are happening in coding and customer support. That's where these AI companies are taking off, getting insane valuations. The desire to make like coding more accessible to everyone. I would love to see that same sort of AI play out in AI, like collaboration, workplace collaboration stuff. So that's what I'm most excited about because there hasn't been that breakout company yet.

34:41And we would like to be one of those. You mentioned, you know, it's been a long journey so far, but it's obviously going well, but it is a grind, right? um and would love to know like how how do you look after yourself how do you get through the grind right how do you get through like you know there's always hard days right you know how how do you manage that well i hear like very popular people in this space that tell companies oh yeah if you're going triple triple double double uh that's not good enough anymore think Think about how many companies are even doing that. Right. So we did that. We exceeded that.

35:24We were happy. So don't always benchmark yourself against like the outliers. Like there's still companies to be built. There's still great opportunities to build in SaaS. You definitely need to be AI enabled and start with AI first lens. There's still incredible businesses that can be built. Like it's not that everything has to be done by open AI. everything has to be done by um anthropic or one of these like sort of crazy companies like if you kind of think about it 11 labs is a great example of they picked one problem text to speech or speech speech to speech and they went really really deep on that and they built like an incredible business that's growing rapidly fast so pick a problem and go deep on it and don't listen to this narrative that, oh, one AI tool is going to solve every problem, right?

36:14So that you need to get because a lot of that is also marketing hype. And there's still incredible businesses to be built today. For us, the best thing for my mental frame is I'm building for the long term. We are profitable. We don't need to go out and raise crazy amounts of cash at crazy high valuations, even if they're offered, which allows us to stay focused on the mission, stay in control of the direction we want to take the company. Like if you're chasing 10 different directions, you end up nowhere. So that's like something really important that we have. And we can navigate this over a longer period of time.

36:52I always believe the best companies aren't just growing the fastest. They're growing consistently for a very long period of time. It's not just about doing that hockey stick and then dying, right? You have to be able to do that for decades. So if we want to build decade-logged companies, AI is something that you can definitely use and ride through. But I don't I don't think it's like, it's to me where, oh, we have to, like, we have to, like, we only have this six month window. And then after that, there's going to be this general AI that's going to take over the world. And there's nothing we can do.