From Startup to Exit
Welcome to the Startup to Exit podcast where we bring you world-class entrepreneurs and VCs to share their hard-earned success stories and secrets. This podcast has been brought to you by TiE Seattle. TiE is a global non-profit that focuses on fostering entrepreneurship. TiE Seattle offers a range of programs including the GoVertical Startup Creation Weekend, TiE Entrepreneur Institute, and the TiE Seattle Angel Network. We encourage you to become a TiE member so you can gain access to these great programs. To become a member, please visit www.Seattle.tie.org.
From Startup to Exit
Gen AI Series - The AI agent redefining travel: Managing Director Mike Fridgen, Madrona Venture Labs
Learn about the founding of Otto from Mike Fridgen, the Managing Director of Madrona Venture Labs where this concept was incubated. Otto is your personal AI travel agent that seeks to be as valuable and insightful as a dedicated personal assistant. Otto offers enterprise-grade travel booking and servicing capabilities through an easy-to-use AI travel agent that autonomously plans, books, and manages your travel, even during your trip.
As Managing Director of Madrona Venture Labs (MVL), Mike Fridgen leads the studio's mission to partner with founders from day one to create extraordinary companies. Mike co-founded multiple MVL studio funds that have created and invested in 30+ companies with aggregate funding of $270m+ and a total valuation of $700m+.
Before MVL, Mike was the founding CEO of Decide, which was acquired by eBay in 2013. At eBay, Mike was the General Manager of the Seattle-area office and the global product leader for professional selling experiences. Prior to that, he was the founding head of product at Farecast, which was acquired by Microsoft in 2008. Since the late 90s, he has been a founding member of three venture-backed consumer technology companies.
Mike is an active advisor and board member of dozens of startups. He is on the LPAC of Pack VC and has served on the University of Washington's Buerk Center for Entrepreneurship board since 2014. He is a graduate of the University of Washington and Harvard Business School.
Brought to you by TiE Seattle
Hosts: Shirish Nadkarni and Gowri Shankar
Producers: Minee Verma and Eesha Jain
YouTube Channel: https://www.youtube.com/@fromstartuptoexitpodcast
Have you know, just like you would with an EA, authorizing the agent to book on your behalf. That's what the product does. And because of advancements in Gen AI and the ability to, you know, pull a calendar data, look at previous trips, previous itineraries, matching up again itineraries with meetings on calendars to know what are those repeat sales meetings, quarterly reviews, what are those meetings that are occurring and predictable, and what are the what are the preferences of users?
SPEAKER_00:Welcome to the Startup to Exit podcast, where we will bring you world-class entrepreneurs and VCs to share their hard-earned success stories and secrets. This podcast has been brought to you by Thai Seattle. We encourage you to become a Thai member so you can gain access to these great programs. To become a member, please visit www.seattle.tai.org.
SPEAKER_04:Hello everybody. Welcome to another edition of our podcast from Startup to Exit. My name is Gary Shankar. I'm on the board of Thai Seattle. This is a Thai Seattle production. I am delighted to have as my co-host Shirish Natkarni, and we have been uh hosting uh this podcast for a year. Uh Shirish is uh author and a serial entrepreneur, but now I should say he's a serial author because he has two books to his name. We shamelessly borrowed the name the title of his first book for our podcast from Startup to Exit. The second book, Winner Take All, is also published. You can buy both the books where books are sold. Um we have been doing the Generative AI series now for a few months, and uh it's very exciting to uh with to talk to our guest today, who Sharish will introduce uh the first AI company born in AI uh in the AI itself. So with that, I'll hand it over to Sharish to introduce our guest.
SPEAKER_02:Thank you, Gavri. I'd like to welcome Mike Fridgen, uh, who's the managing director at Madonna Venture Labs. I've known Mike for a number of years. Uh we have partnered with Madonna Venture Labs and Thai Seattle to offer the Go Vertical Startup Creation Weekend, uh which we ran for about seven years or so, uh, incubating you know numerous new ideas. Today our discussion will be around a new um generative AI application that MVL is launching called Auto. Uh but before that, uh I'd like to quickly introduce uh Mike. Uh Mike is a uh longtime veteran of uh of Seattle. Um he was the CEO of DeSite.com shopping decision engine that was acquired by eBay. And then uh he was the CEO of Faircast, uh one of the first uh travel services out there that allowed uh um people to figure out which directions fares were headed so they could make an informed decision as to which uh whether to buy the ticket at that point or to defer the decision to a later date. Uh and then he um became the managing director of Madonna Venture Labs, uh, which was originally a spin-out from Madonna Ventures, uh, but now has many other investors as well. Uh, and they um incubate uh new companies and spin them out as they're doing with this company Auto. So welcome, Mike.
SPEAKER_01:Sharish, thank you so much for having me, Gary. Really appreciate uh being a part of this. That's great.
SPEAKER_02:So uh before we get started, maybe you could uh tell us a little bit more about um how you uh you were an entrepreneur first and then decided to join Madrona Venture Labs to incubate ideas typically from other entrepreneurs. So what made you decide to make that shift and uh and how is it going so far?
SPEAKER_01:Sure. So a little background. I'm a Seattle native. Uh started my entrepreneurial journey uh as an undergrad at the University of Washington, started my first company in the late 90s. It was a travel, student travel technology company. And it was really at the intersection of two major trends, this first company called Triphub. One was students were the largest demographic on the web at the time. And two, travel was the leading e-commerce category. So you kind of looked at those forces coming together and launched this company. Uh, it was a company backed by Madrona uh and other investors, uh, and that that company uh went on to be acquired. From there, uh, some twists and turns, but ended up back at another startup on the founding team of a company called Faircast that you mentioned. Hugh Crean was the CEO of that company, and it was exciting to join him. I was the product leader of the business. But we were doing some really interesting things. Um, Oren Ezzioni, who's a computer science professor at the University of Washington at the time, he went on uh to be the CEO and really the founding driver of AI too, uh, and now a new company, True Media. So a real legend in the world of AI and entrepreneurship in the Pacific Northwest. But it was his initial technology and and and the work of many others that helped build a price prediction engine and travel search engine that was Faircast. Company was acquired by Microsoft after a couple years there. Then I went on, as you mentioned, to decide. This was taking some of the learnings around using data to inform better decision making and applying it to a much broader shopping category. Uh, and so we were helping consumers know what to buy, when to buy, ingesting massive amounts of review and pricing data. As you said, we sold that company to eBay. Uh, after a couple years of helping build up the local uh office here in the in the region, it was in Bellevue, uh, got the call from Madrona to say, hey, you know, we want to invest more and really build on an initial start in the Madrona Venture Studio, which was an incubation arm for Madrona. And that's when I came in. So I've now been at Madrona Venture Labs for nine years. Uh we've built dozens of companies that have gone on to raise hundreds of millions in funding. Um, many of these companies are are are continue to grow and and um right here in our region. And and so it's been it's been amazing to be a part of that with a great team. Again, some of which you've had on your podcast, Jay Barteau in particular, who uh is the the leading technologist on the team and responsible for a lot of the innovation coming out of our studio. So that's you know, if you think about, you know, three one half of one chapter of the career was around founding, building as an entrepreneur and a founder. And the second chapter has been supporting founders, um, supercharging their efforts to help them be successful. And in terms of you know, the long-standing relationship with Madrona, it was really a spend 25 plus years now working with Madrona in some way. So they were investors in those three startups I mentioned, and then of course, the at the core at the heart of what we're doing here in the venture studio. That's great.
SPEAKER_02:So, um, how does the model work at um MVL? Um uh do you um incubate your own ideas or do you uh typically like to partner with founders to incubate their ideas, or it's a combination of both?
SPEAKER_01:Yeah, we're all about working with the most ambitious driven talented founders. Uh and so we'll meet them wherever they are in that zero-to-one process. And whether you know they're coming up with the idea and we're supercharging their efforts, or we have an idea and we meet them along the way, uh, and they come in and co-build with us through a process. Um, you know, our reason for being is de-risking and fast tracking their success. Um, you know, we know as founder builders and in our community, we know it's all about uh a founder's ability to take an idea and learn and iterate, uh, out-execute the competition. And so that's gonna take uh a lot of different um, you know, iterations and and and cycles to get there. And so we really are all about putting the founders in the best possible position to have success and uh and surround them with the right uh resources and support to get there. So, you know, we work, we have three different models for how we work with founders today. Um, in some cases, we're co-building from the ground up. We call this our core model. It's really we're really built to be incubating ideas and growing companies from from the ground up. Again, sometimes our idea, sometimes their idea. Other times it's more acceleration. We meet a team that has some progress and we'll come around them and partner to move the company forward. And then in other cases, it's purely an investment. It's a it's a small investment, generally the first money in, and it comes with uh us as a team advising and supporting the company um you know as it as it first gets going. But everything's very early, zero to one, formation, precede. Madrona has a core fund that's seed and A, and then they have a growth fund beyond that. So uh there's a phrase we use that that's used at Madrona, day one and for the long run commitment to founders. We're that very early formation preced stages of company creation. That's great.
SPEAKER_02:So let's talk about um auto. Um it's a very interesting idea. Um uh tell us uh first what auto is, and then um how did the idea come about?
SPEAKER_01:Yeah, sure. So auto came about it. You know, we have a practice at MBL of every off-site, everyone needs to bring ideas to pitch. So um, you know, of course, you know, we look at the portfolio and we do all the things you'd imagine a venture studio would do at an off-site, but a big part of it is living the same thing that the founders that we're meeting with are doing. So we pitch each other ideas, and and about a year ago, auto was pitched at an off-site. Uh, and then very quickly after it was pitched, because there was so momentum and the team rated it really strong in terms of thematically aligned, uh, the right time to address it, and leveraging our strengths as a team. We have I mentioned my background, a couple travel startups, and Madrona has a clear history in travel, and Steve Singh has led multiple bets in that space, Spot Nana, most recently direct travel, troop. So we had a lot of domain depth as a venture firm and as a venture studio in that. So it made sense for us to you know do our work and validate and do our diligence to see if there was something there. So from there we started talking to experts. Uh, we talked to over 50 people in the industry to you know understand, you know, why had previous attempts at an AI travel agent not work. I mean, there were a handful of companies that raised significant funding about a decade ago, and they just really never caught traction. So we just did a lot of a lot of work to learn. But specifically, what the idea is, is um auto is an AI travel agent for business. Um, and specifically looking at the unmanaged business segment. So in the travel sector, you have leisure, you have business, and then within business, you have corporate programs that are specific platforms that users use to book their travel. But a big percentage, almost about half of that market books their own travel. So we're looking to give them essentially an EA, a travel agent who will plan and book all of their business travel. So that's what the that's what the product does. And because of advancements in Gen AI and the ability to, you know, pull a calendar data, look at previous trips, previous itineraries, matching up again itineraries with meetings on calendars to know what are those repeat sales meetings, quarterly reviews, what are those meetings that are occurring and predictable? And what are the what are the preferences of users? These airlines, these hotels within walking distance to the office, all these specifics, and get really smart about um pre-planning, um, having, you know, just like you would with an EA, authorizing the agent to book on your behalf and really going through all those steps and then leading to the moment of travel, where of course there are all the kind of hiccups we've all experienced flight delayed, um, you want a room upgrade, all these things that this AI agent can do in an automated way on your behalf to ensure you have the perfect trip. So that's what uh that's what auto is going after, um, really providing you that expert travel agent who's doing all of the work to be proactive and automate those tasks that today, if you kind of think about it, we're still operating like we were decades ago. We're going to booking engines, we're entering all the same information, then we're going through transactional flows. These are tired processes, tired experiences, and and this is a moment to reimagine that.
SPEAKER_02:Right. Um, so you did some uh validation, as you said, talking to experts. Um, did you do any validation with the actual end users themselves? Uh absolutely how comfortable they would feel. Um because you know, you know, I plan my own, I'm not a business traveler anymore. I'm semi-retired, but I plan my own travel and I like to look at the options and make decisions myself, etc. Um, you know, it may be hard for me to trust um an agent, an AI agent to do the right thing uh for me. So was there um enough interest, obviously, from the uh end users?
SPEAKER_01:Well, it's such a big part of what we do. I mean, I'll just back up a minute just to kind of talk about our validation process. So, with any idea that we put through our process, you know, most ideas are dying along the way in the validation process. We have three phases validation, traction, building the product, and then fundraising. So within validation, there's three components market, customer, investor. You're really hitting on you know that customer component of what we do. And we spend a lot of time here, and it's probably the most important thing we do. Um, and really where again, most ideas go to die because you know, very quickly we aren't seeing the that that outlier signal, that pull from the customer that this is a major problem. These tired booking flows going to the booking engine that you were at just last week and it doesn't remember anything about you, these same tired flows of going through it. We heard a lot of pain around both planning uh and and buying, but also in-trip experience, making phone calls to change reservations, all those issues really arose through the conversation. Um, and so, you know, both experts and customer validated with clear signal that there was an opening in the market to address this. So many people are unhappy with you know their their business travel booking that they bypass, they we call it in the industry going rogue, they bypass the existing corporate tool because it's so painful to work through. So there's there's a growing number of business travelers that are independent, booking their own travel, but then they're adding all that workload onto their own plates. And in an age of you know, this world where I think we see in the future where AI agents are completing these vertical tasks for us, it's the perfect scenario.
SPEAKER_05:Right.
SPEAKER_01:Uh, because it is such an antiquated process, and we can to a level of confidence and uh and and and and high value uh perform these tasks. Right.
SPEAKER_02:Now, why are you limiting yourself to um one segment of even the business segment? Why not? I mean, you know, take my example, I'm semi-retired, I'm traveling a lot. I'd love to use a tool like this. Um, I you know develop some trust and all that. I'd love to use uh so I mean expedia is not necessarily I they have, of course, a business section, etc., but it's a tool for everyone.
SPEAKER_01:So why can't it be tool for everyone? Right. Yeah, I mean, a couple of reasons. One, the business travel sector and in that segment of travel behaves differently. Uh it's predictable, it's repeatable. Again, you think in our in a in a hybrid and remote world, people are coming back to HQ quarterly for certain types of meetings. People when you know on their calendar are inputting their itineraries, their flights or hotel information. It's all tied to a key meeting that's on the calendar. A lot of that data then we can ingest, and we can start to build real memory around that and an understanding of not only you and your preferences, but your travel patterns and behaviors. So that's and it's also for a more um transactional kind of experience, even for the user. You think about leisure travel, and there's a lot of joy in planning a family vacation to Italy and all the things that go in session after session. It's like 40 websites searched, I think, for for the average family vacation before you're making a purchase. And those hope happens over weeks and months. Not true if you're going to New York for your fifth sales meeting of the of the of the year. You're staying in a similar hotel, you know, you're you're taking similar flights where you have your loyalty programs. It's just a different mindset and it's a different motion. It's the right place for us to start. To be sure, once we really know consumers and their behaviors, and you know, we can, you know, there's all kinds of ideas we have to moving into leisure overtime. But we're gonna start with a segment that I think really benefits from our ability to uh bring that data in and apply it uh in this multi-agent system that we have envisioned.
SPEAKER_02:That's great. So now uh in terms of how you build this technology, I read somewhere that um um the generative AI model that you use primarily was uh GPT-4.0. Um I don't know if that's accurate or not. Uh can you talk about uh the kind of the architecture uh of the solution and how it was built?
SPEAKER_01:You bet. Yeah. I mean, first, to this is a very modern AI application. Uh so I think in a world where uh apps will be built in a much different way, uh you know, where there it's multi, it's it's multiple agents playing a role in an orchestrated system to accomplish a complex task. So in our world, we are using large language models uh tied to different data sources. To give you an example, we have one agent that plays a role around calendar. It ingests calendar, as I said, itineraries, meetings, and it understands and makes sense of that to understand your preferences and your behaviors. We have another agent, which is an LLM tied to a travel API that's pulling back real pricing and availability for flights and hotels and other and other travel components. And you can imagine multiple agents who are then all play different roles and have different prompts tied to them. We also have a personal biased agent that understands not only long-term memory, preferences of the past, but preferences and intent you've shown in the moment of indicating a trip or of directing a trip. So all these systems are coming together. And in terms of the models we're using, there's a lot of flexibility here. You know, we'll have different models for different purposes. Um, in some cases, we might have a performance objective, other cases, a cost objective. Other cases, we might need you know a stronger model with more capability around you know, understanding uh, again, these these personalized data sets and being stronger in terms of how we articulate what we know about a user. So we have flexibility in what models we plug in for different use cases, and that'll really evolve over time. I see it being uh multi-model, though. This is not a single model we'll use on the back. This will be a set of models that all play different roles in the system.
SPEAKER_02:And those models may change one day you may be using you know GPT-4.0, another day you may be using Llama or something, depending on how you need to change.
SPEAKER_01:And almost certainly with the rate of change, uh that's almost certainly uh the only thing we know is that the exact set of models we're using today will will evolve and change. That's great.
SPEAKER_02:All right, uh, let me turn it over to Gaudi to continue the discussion. Over to you, Gaudi.
SPEAKER_04:Uh Mike, it's uh interesting. You you kind of said uh you know, modern AI uh company. I I thought you were an AI-born company. But you already there's already a the three years that models have been around, there's already a you know AI. Um we had to modernize how we think about uh language, you know, ML and AI models. It's very interesting. So let me start with this premise, right? Post COVID, would you say the segment you're going after has a pro-leisure uh approach to travel, meaning I go to headquarters, but I also extend it two days, or I go to someplace I extend it two, three days, and or stay two, three days to meet with my other colleagues around a meeting. So it's somewhat not it used to be you left on a Monday, came back on a Wednesday, but then your leisure trip started on a Friday. Seems like that's blending. Would you guys consider that?
SPEAKER_01:There's no question it's blending. There's a term in the industry, it's called bleisure, a mix of business and leisure, and that's very common today. Uh, and so you know, those are things. We're thinking about in the experience and the interaction model of planning and booking these trips. But yeah, I mean, you know, you were speaking a bit about modern AI and modern apps. I mean, we are, you know, inventing as we go here on the edge of what, you know, as we're learning and and iterating on what what what these what's possible with this technology. You know, some people think we'll have an agent, uh, everyone will have a personal agent, right? It'll help us accomplish tasks in our lives. Um, almost like everyone has this supercharged EA in their life, not just for professional reasons, but for personal reasons. And if you but when you think about these specific verticals, they're so complex. Uh and they they tap into deeper data sources that need real time around. I mean, flights is in and hotels are a great example. I mean, those, those that data is not sitting in the models today. That's real pricing availability. It's pull down in the moment, that changes in the moment. And so these more complex systems just require um uh a more a more uh focused vertical execution. So I you know you can imagine that there'll be an orchestrator personal agent to every human, but underneath that are a set of vertical agents that have deeper capabilities.
SPEAKER_04:So let me sort of kind of uh build on that, right? So obviously there's co-pilot from Microsoft for everything at the moment, or at least their offerings, right? Uh you'd have to assume that the segment you're going after also has an overlap with a copilot, right? They understand already a lot of things, meetings, calendars, travel patterns, etc. Um, could they not extend their own models or their own understanding and um offer it? Or do you do you see them not playing really vertical and instead being the horizontal layer partnering with somebody like you?
SPEAKER_01:Travel's a really uh fun use case, and a lot of companies, large companies, use travel as their demos.
SPEAKER_04:Yeah.
SPEAKER_01:What we find though, beyond the beyond the demo, beyond the big event, it's very tough to pull these off uh the complete end-to-end, the planning, the transaction, all the pieces that go into making this work. So I absolutely believe uh it'll take more than what you see today to pull this off from from the large travel, or excuse me, the large tech players who are who are running these really fun demos. I think if you want to uh build an itinerary, sure. Um you want to plan a vacation and and have one of these LLMs ride alongside you in a co-pilot style to help you plan, sure, that's fine. But when it comes to all the way through to booking a complex, multi-destination, uh multiple leg family trip to Italy, co-pilots aren't there yet. And it'll be a while. And so that's the type of depth that we're going into at auto. Again, not focused on leisure initially, but uh within this business travel segment, get that right, refine it, then scale it and and and bring it to more scenarios within travel.
SPEAKER_04:So uh just last question on the on the competitive thing, right? So let's just take two specific Google and Expedia. They clearly are behemoths or legacy behemoths in in this space. Uh Google's got their own model. Expedia, I think, has announced some assistant of some kind. Uh uh, so how did you guys thread the needle saying, are they not addressing this uh particularly, or because they have Expedia being leisure-centric and Google being just price-centric or flight, you know, data-centric? How did you thread this uh to pick this segment, the business segment? Because they have the data, they have behavior patterns over time and preferences, right? Uh etc. Why why the should I depart and come to you?
SPEAKER_01:If you look at online travel agencies, speedia, there's others, they've had the data for decades.
SPEAKER_04:Yeah.
SPEAKER_01:If you if you go back in the way back machine, go back go back 20 years, you're gonna see the same experience. It's the same book, origination, destination, dates. Same funnel.
SPEAKER_04:Right.
SPEAKER_01:Most of them still don't know you, and you've used you've used them dozens and dozens of times right over decades.
unknown:Right.
SPEAKER_01:They hardly know you. They've they've had the data. We don't see personalization, we don't see innovation.
SPEAKER_04:Right.
SPEAKER_01:So uh for any startup there that's worried about that, just go execute. Out execute. I mean that it seems pretty clear. If you look at Google and you look at the major search engines, there's a there's an innovative dilemma issue there around business model. Yeah, they don't make they don't make money by helping you um with an efficient um automated transaction. They make money by sending you to multiple places, right? Multiple travel websites. It's an ad-driven. Same with the travel search engines, the kayaks and the rest. They don't make money by managing end-to-end the planning and the transaction piece.
SPEAKER_04:Got that's right.
SPEAKER_01:So you disruptive to their business model. And so, how quick will they be to innovate and and and bring new travel experiences to market? Uh again, I think we see the same referral model there uh with with Google Travel today. And and and other, not just Google. I mean all all search engines, it's a it's a challenge.
SPEAKER_04:So do you then see all the say the the partners, airlines, hotels, all the players who actually um provide these services to your uh to your customers as your partners is do you see I mean they they have this uh sort of an interesting relationship with OTAs? Like if I if you're Hilton, you have an interesting relationship with OTA because you have or Google for them, because you're spending so much to attract your customer while here you're delivering a customer, right? Because you're saying, hey, I want to I've stayed in Hilton the whole time. I'm uh whatever, Hilton gold. So I'm gonna steer you to Hilton. How does this relationship of inventory owners and auto mix? Or are you uh going to let them come to you? You're gonna scrape them? How how have you guys thought of that?
SPEAKER_01:We have access to travel APIs, it'll give access to pricing and availability. You know, I think a major focus for suppliers, airlines, hotels, others are to drive direct bookings. They've invested in more efficient APIs of their own. Um they'll create incentives uh and better user experiences for to attract uh customers to book direct. There's a there's been again, this has been decades movement to drive more marketings directly through suppliers. And and we'll and we'll do a mix of of dropping direct likely and and pulling from travel APIs uh and and and booking through those sources. So it'll be you know, we'll it'll we'll optimize over time.
SPEAKER_04:So, what is then the business model?
SPEAKER_01:I will say too, this is what's great about you know the founding CEO of Otto is Michael Goleman. Uh background uh in in the travel industry, had worked at Expedia on a uh chief product officer role for both consumer and on the business side, uh, deep network and understanding. Also, Steve Singh uh is a very important part of this team. Um, he's the exec chairman of the company. Um, he's also leads investments like Spot Nana and Troop and Direct Travel. So you think about all these pieces together from a data and distribution standpoint, there's some real advantages to the the travel depth and the travel you know network um that I think will will definitely be unique for auto as it moves forward.
SPEAKER_04:Right. So what would be the business model or what's the business model at the outset? Maybe it evolves, but at the outset, what would you because unlikely you can get everybody to pay you just to they they because they have a process that quote unquote is free, although they're trading their time in, and so it appears free. But what's the business model you guys are thinking about now?
SPEAKER_01:I mean, some of the things you could imagine, I mean, this early, uh, and and Michael, the CEO, is going to is set this strategy and make these decisions over time. But what you can imagine is you know, opportunities for subscription. You can uh you can see opportunities for transactional revenue. Um, so there this is a company that can generate revenue day one uh as it enters the market.
SPEAKER_04:Right, right. You know what's interesting about this particular application is it actually is the connective tissue, right? I mean I'm I'm a power traveler, so the bleisure kind, right? That travels for business, takes advantage of the leisure. And I know my preferences, right? And um let's assume eight preferences too, right?
SPEAKER_01:You've done you've done it enough that that's all sitting there. Yeah, yeah, yeah. Your email. That's all some that's all information that can be gleaned. And and LM such a nice job of summarizing that back to you in a natural language, in a way that sounds like, wow, this is a very educated system telling me how I travel. It's actually kind of amazing how well it does that. So I'm with you.
SPEAKER_04:Yeah, yeah. So and and the thing that frustrates frustrating is now, if you take my last three trips, uh saying the last two months, right? The patterns are not that different. They're similar, right? Uh, I decide to go, I park in the same spot, I have the same amount of time before my flight. When I enter, you know, my location is declared. I use uh, you know, and then I use the same club in the in the airport, mostly the same airline. Uh, and you could see from my patterns that my preferences are non-stops as opposed to you know, jump around, you know, uh uh etc. And I have preferences of airlines, hotels, uh uh, you know, prefer uh uh share ride sharing apps over rental. I mean, it there's a pattern that's emerging. Almost every time I have to repeat it every single time, uh, you know, start from scratch to go through this all. Even if I assume that I can get past that, when the thing breaks, that is flights delayed, uh, something happened, there's no, then I have to start over again. And that's the part that's missing.
SPEAKER_01:Uh so right. I mean, we hear we hear that in our customer interviews. Uh flight delay, but also meaning changes. I mean, oftentimes in a professional, you know, you know, cadence that kind of goes through week meetings change, then all the travel plans need to change. Now you have to go back in there and unwind them and rebook them. What a pain. Why are we doing that? No one should be doing that anymore. We should not be going through these tired transactional booking engine flows. It's it's amazing. I mean, you know, you kind of the the flip of the green screen. I mean, that was empowering, right? A human agent would do all the work for us. They flipped it over, then it gave us the power to do it. But now we're doing all that work. Hours and hours, you know, planning trips.
SPEAKER_05:Yeah.
SPEAKER_01:And then the meeting gets canceled, and they're replaying you know, it's insane. So the world needs a better experience, and we're at a moment in time from a technology standpoint to give it a fresh new look. And that's what Michael and the team at Ottawa are doing.
SPEAKER_04:At the time we are recording this, Hurricane Milton's uh threatening Florida. I'm just imagining the amount of business and whatever um uh conference travel that's going to get canceled in the next two weeks because you know, Florida, the greatest destination for conference travel, is going to get canceled. I can't even imagine the number of things that that uh that's going to change. Do your partners see uh, I mean, when you pick this, right? You picked it from the user perspective, right? So now there's also the partners perspective. We're all investing in some AI, etc. Do they see the value when you talk to them? Did they see, hey, we need somebody like this so we could have a better interaction with me, the traveler? Because that that's sort of the worst part at the moment. I mean, whether I call the airline or whether I call it's always like you're in line. They they have no, they can't tell the difference between me who's traveled for them 20 times versus somebody who traveled for the first time. So it just can't do it. What do they see as an advantage working with you guys?
SPEAKER_01:Well, it depends on what partners we're talking about, right? So right now we're taking a user who travels for business, but they do their own booking. So they're going directly to the online agencies or the suppliers' websites to do the work.
SPEAKER_05:Sure.
SPEAKER_01:So they're all a partner essentially, right? And we are using different methods to access that price and availability and complete the whole loop for them. Sure. Both the planning of here's like you said, you have repeat behaviors. We're tapping into that to make that planning efficient, and then we're closing the loop, we're doing the transaction. You know, the user is literally authorizing, like you would a human EA to do the booking and then providing all the information in your calendar and the itinerary nice and clean, right to you. So the partner is the user. Now, in a world in the future, where we start to blend kind of into the corporate side, then you could ask, hey, for corporate travel uh TMCs, uh you know, travel management companies, is there an opportunity to partner there to make their experience more elegant, interacting with their users? That's interesting. Over time, we might go up market into more corporate. You talked about leisure and you know, moving down cross-market into leisure travel. So I think we've got the right entry point. Often with startups, you find that wedge where you you believe you can create an incredible experience and really gain momentum and habitual use with a customer, a high-frequency business uh traveler who's traveling multiple times per month. They they kind of live in these systems today, right? And we're gonna take that all off their hand, make it make it faster, smarter, and a and a great trip, a really great trip experience. But over time, we can we can move into these other markets uh as the technology improves and we learn more from our customer interactions.
SPEAKER_04:Great. So let me sort of um go up a level, right? You put on your MVL hat. So you said, hey, we listened to a lot of pitches within your team from outside, etc. So our audience, a lot of entrepreneurs out there, I mean, the democratization of AI, therefore they access the LLMs, etc. Um, now lots of opportunities have are coming about that you can go start companies. So what would you advise entrepreneurs to go look for? Uh other than the basics, right? Validation of idea. How do you find spaces uh that you have found which from the outside looks very crowded? Hey, Expedia is going to do that for sure, or I'm using them as an example. Uh you know what I mean? How would you uh uh advise an entrepreneur to say, AI, when you look at AI opportunities, how should they view uh the opportunity itself as though they're coming to MVL to say, I have this, and if they were to pitch to you, what are the things you want them to look for before they even come to you so they can eliminate bad ideas?
SPEAKER_01:Well, I'll tell you something, it's compelling. It's compelling when the founder comes to the to the pitch with unique insight about a problem. Maybe they've lived at themselves as a consumer over a long period of time, maybe they've lived in that industry and and they've had unique access to problems and insights derived from from being in the domain. So, you know, oftentimes we talk about the importance of founder idea fit. But what comes from that that's unique and interesting is those insights. Um, I mean, when we're thinking about things we worked on, we just talked about the depth and travel as a firm in Madrona and within our studio and as a people, the experiences we've had to give us unique insight into that. And that's pretty consistent. I mean, if I think about you know what we're looking to do, we're looking at where does Seattle have strong talent? Where do we as operators have unique insights? Where does Madrona have an investment track record, right? They have, you know, judgment and and they've made decisions and learned in these segments. And we're we're taking all of that to help hone our focus to give us an edge. Because, you know, there's a lot of things you can do in this moment, and it's pretty quick to whip up a demo, a Gen AI demo. But really, you know, what's your story for you know, how are you going to create an excellent product? Um, what's your unique insight into the problem so you're positioning yourselves in a unique way? Those are the things that help founders really uh stand out and and kind of become those outlier founders who, you know, uh I think investors are excited to back.
SPEAKER_04:So it's it's great to hear that the the sort of the back to basics moment, right? Hey, what's your idea? Why you should do it? Do you are you the right person team to do it? Um how are you going to go about it? What is it that you will bring that others can't bring? I mean, the these processes or these set of things are not unique to AI. Was there along the way? Uh glad to hear that, you know, because there's so many demos you see in Gen AI at the moment. All of them are compelling, but none of them are unclear. They are companies, you know, so to speak.
SPEAKER_01:Yeah, I think in this moment, I mean, you just hit on it, Gabriel. I think this understanding the role that that LLMs play in solving the problem in a unique way that wasn't possible before. Um, there is a lot of hype. There are a lot of, you know, AI is sewn around pretty casually in those pitches, those conversations, breaking it down into okay, here's the scenario, here's here's what LLMs do really well uh and can do consistently and at a production level and scale level. Here's here's what in our our product where it really plays a specific role. Because with us, like in this travel scenario we're talking about, it's not the whole product. It plays a role. It is not the whole product. There are a lot of components to this. Now, it unlocks some things that weren't possible before, uh, around summarizing your profile, um, storing that memory, applying it to future. There's things it does that are really important that make it as the whole thing stick together in a modern, innovative experience we haven't seen before. But it's not, it's not, you know, we we built something on top of a core model, and we, you know, that's there's so much more detail and nuance to the execution of the architecture and the product under the hood. And so I think having a deep understanding of that and be able to articulate that is really important in in connecting the dots. Why, you know, as a founder, why me? Why this? But also, what role does AI play on unlocking these new capabilities that that customers really care about and wanting to pay for?
SPEAKER_04:Got it. Great. Sharish uh back to you. This is uh Mike, fascinating to hear from uh uh from a longtime founder. And uh and you, you know, having started essentially in your dorm room, the first travel startup it looks like, right?
SPEAKER_03:Today you're back uh full circle doing another one of those. And there's somebody, some other Mike in a dorm room at you now thinking of this, baby. But uh it is fascinating to think that you stayed stayed the course the whole time.
SPEAKER_04:Uh incredible, incredible. So, Shirish, back to you.
SPEAKER_02:Thank you. Fascinating conversation. Uh, I guess one last question was uh uh when do you expect uh auto to launch? Uh when can we look forward to trying it out?
SPEAKER_01:Yeah, I'll I'll leave that to Michael uh Goleman, the CEO, to announce. Uh he's done incredible work, he's building the team, building the product. Um, you know, the team is is iterating on it every single day. Uh new builds, new capabilities. Um, so I'll I'll leave that to Michael to announce.
SPEAKER_02:Okay, very good.
SPEAKER_04:Excellent.
SPEAKER_02:All right, thanks uh again, Mike. Uh it's delightful to uh talk to you today about auto. I think uh it will have the same uh impact on the travel industry that Expedia had when it first launched, whatever, 25 years ago or more. Um so look forward to trying it out uh when it comes out and uh you know happy to bring other ideas to MVL in the future as well.
unknown:Thank you.
SPEAKER_01:With appreciation having you today. Yeah, thank you.
SPEAKER_04:Thank you for listening to our podcast from Startup Exit brought to you by Dai Seattle. Assisting in production today are Isha Jen and Mini Verba. Please subscribe to our podcast and rate our podcast wherever you listen to them. Hope you enjoyed it.
