The Exchanges

Q Okay. And that was gonna be my next question is, um, is it, you mentioned earlier software, is it exclusively, uh, software or will you include hardware as well?

A Uh, you know, earlier when we started the, the fund in 2011, we did, uh, invest in a few hardware plays, but over the years we realized to invest in a hardware play, A, you need a longer term perspective, B, you need, uh, a lot of capital, and C, uh, the sales cycle and adoption takes a lot longer. Uh, so we are a early stage, uh, micro fund, Uh, it is not a right strategy for us to be part of, so we've sort of walked away from that, and we exclusively now invest in software. They could be some hardware component where it is a data gathering tool or something of that sort, but the idea is the power is around, um, you know, the data that we are collecting or, uh, collecting, or it could be the decision layer That it's enabling the physical environment it is in.

AI assessment note: “so we've sort of walked away from that, and we exclusively now invest in software.”

Q Wow. So, uh, how long ago did you start the company and, and what's the, what's the evolution look like since inception?

A Yeah, we started in 2019, um, January of 2019, um, we grew about 50% a month, every single month from January of 2019 all the way up to March of 2020. Yeah, I don't have to tell you what happened in March of 2020, um, but essentially we dropped like 90, 80, 90% almost overnight. Um, and then pulled ourself out, uh, took us about a year and a half, two years, pulled ourself out, uh, and, you know, now we've got, you know, thousands of businesses on our platform, you know, uh, you know, millions of consumers on our platform, and it's, um, you know, been, been a, a fun run, but, uh, uh, but definitely has its ups and downs, and, and we, you know, rolled those downs, learned from them to get better at the ups.

AI assessment note: “we started in 2019, um, January of 2019, um, we grew about 50% a month”

Q And then did you intentionally or make a point of getting, getting everyone together throughout the year?

A We did. We had, uh, we, we, we joked, we had onsites instead of offsites. We'd have onsites. Uh, we would fly. So I had people in India, um, uh, Europe, California, Atlanta, South Carolina. I'm here in New York, someone in Denver, someone in Chicago. I'm sure there's a couple of other spots. Uh, I picked Atlanta actually as my, as our, as our hub. Because it's everyone, there's always a direct flight to Atlanta and it doesn't have the same overhead as New York does. So what we did about once a quarter is we, we find a large Airbnb, like a 10 bedroom house or whatever. And, uh, we'd hole up there for four or five days. We would do some productivity sessions, some culture building sessions, some volunteering. I always would have a volunteering session in the midst of what we do. Uh, we would cook all our meals. I mean, we'd go out once or twice, but we would do a thing about cooking together as a team. And we'd always do a sort of a variation on a hackathon. I would call it, I always called it a scrimmage with the subtle difference between the hackathon where the goal is just to sort of prototype. The goal was to actually ship. So we would pick some very small thing. It could have been just a feature with an Augie, but a couple of times it was standalone products. Uh, and, and build and ship them inside of 48 hours.

AI assessment note: “We did. We had, uh, we, we, we joked, we had onsites”

Q And so how does that, so you get 3800 some odd applications. How does that translate into the selection committee making decisions as to which founders to bring into a batch?

A Right. So we use humans so far, and we'll continue to use humans. I'm sure there'll be some AI in there, but so as I mentioned, you know, we have alums, Berkeley alums, who are very excited about helping us and other people, not just Berkeley alums, but anybody who goes, oh, this is, this is a cool place to be. I like helping startups. So we have about 900 people Mostly Berkeley alums, but not all who have signed an agreement to be advisors for our startups pro bono, no charge. We don't pay them and they don't charge the startups anything when the startups are in the program. So of those 900, a few hundred are part of our selection committee. So we divide up all those applications by industries. We assign them to advisors based on their expertise and we say help us score them, uh, online. And so from that big 3800, we narrow that down to about 170 for a first round interview. And again, we invite advisors based on expertise. Uh, sometimes we'll invite Berkeley faculty if it's highly technical and we really need some deep technical expertise. And then from that, we narrow that down to a set of finalists of about 70. And that's a longer interview with even more due diligence. And voila. 20 or so companies, um, are selected for the accelerator batch.

AI assessment note: “from that big 3800, we narrow that down to about 170 for a first”

Q Wow. Um, what is, uh, what is international response been to the program?

A It's been huge. So we are a university program that doesn't teach or do research. So like any one of those programs on campus, we have to raise our own funds. So we do that primarily through international partnerships. So our main source of revenue to support the program is partnerships with mainly international governments. So our biggest partner is JETRO, the Japanese External Trade Association. They send startups to participate in our program. They don't get funding. They're not accelerated, but we have a special program for them. And they pay us a sponsorship fee per startup. And then more and more we were being asked to go to other countries to do boot camps and workshops and bring our learnings to the startups in their home country.

AI assessment note: “It's been huge. So we do that primarily through international partnerships.”

Q know, do a really succinct and clear job of helping farmers understand what the return on investment was going to be? And if so, you know, did, did you ever find that it was nebulous or did you figure out how to crack that nut to make it clear to farmers that no, you know, employing our sensors or actually either, you know, generate top line or improve your margins?

A Yeah. Yeah. So that's why we focus on very specific problems on the farm. It's problems where in your head, you can actually calculate the ROI of deploying a sensor solution. So that's why we started with leaks, maple tubes in maple, because they know that they spent 18 hours a day, dozens of people, Uh, finding those leaks so they can visualize, Hey, I'm going to spend like a quarter of my time that I used to finding those leaks. Um, you know, it's the same in California with, with theft. It's, Hey, every time this happens, you know, it's a 40% chance every year that this happens, you know, it's a 50,000 dollar Damage every time it does happen. And so it was like, oh, Hey, yeah, that, that's a, that makes sense. Let's do that. And so those are the problems we try to enter with. And then, you know, there's more of the crop up that are less hair on fire, but that they can expand with.

AI assessment note: “we focus on very specific problems on the farm... where in your head, you can actually calculate”

Q So, uh, if we can, if we can go back there, and also, what did your, your career progression look like from the start, and, and how did it get to Free Will, and ultimately Tidal Wave?

A Um, okay, so it has to start from many years ago. At this point, I think it's like almost 30 years now. Uh, I started my career as an engineer, and I still consider myself an engineer today. Um, so I started my career as an entry-level engineer at DoubleClick, which is a internet one point no company. Okay. Uh, and then when, you know, I rise up I rose up to light the, ah, engineering team. Sold to Google. That's when I left, started my own company called Freewheel. And to build a different capability, still in advertising, but build a different capability to power TV advertising. Uh, and then seven years in Comcast made the acquisition. I stayed at Comcast for another seven years. They helped them to acquire a company, five of them, and putting all the platform together for, you know, basically TV advertising from buy to sell. Okay. It's a exchange. That's when I was like, you know what? I've spent 20 plus years in advertising. I want to learn something new. So I tried something very different. I jumped right into FinTech. Add better mortgage as their CTO. Learn everything about mortgage origination from beginning all the way to the end. And then realize as a lender, the reason that all those lenders are facing such a big challenge is because they don't have the right tool. So I left the better with the mission. I want to build a capability as the technical platform utilizing …

AI assessment note: “I started my career as an entry-level engineer at DoubleClick”

Q Makes good sense. Um, you're in Michigan. Uh, I, I'd say it's fair or hazard to guess that, uh, most of our audience, myself included, thinks of Motor City and in the automotive industry. Um, has there been, um, you know, a, a, a focus on automotive, uh, over the years, or is it more broadly encompassed, uh, more of the state's, uh, industry?

A Depends where you're at in the state. Detroit is still the automotive hub, and in fact, um, there is a pretty large, uh, sort of co-working space accelerator in Detroit, uh, called New Lab that has just taken the place by storm. It only opened two years ago. I think right now there's about a 125 startups in it, and companies moving from all over the world to be near their customers. Uh, and there's lots of other activity going on, principally in the Detroit area, related to mobility, and I'm feeling pretty optimistic about Advances in mobility sort of broadly defined within Detroit. Uh, where our, our office is in Ann Arbor, um, it's the University of Michigan, the largest public research university in the country, doing two billion of research a year in everything. In Ann Arbor, I would say, in particular, their strengths in cybersecurity, in biopharma, um, in, um, supply chain, In battery technology, and, and several others. I mean, one of the, the benefits is, with a large university like that, it's good at just about everything, and there's a lot of research coming out. Go up the road, Michigan State University, another large research university, probably one of the two or three leaders in the nation in agricultural technology. Um, and so, we've got that sort of in the center of the state. You get on the western side of the state, um, and there's, A lot of, there are a lot …

AI assessment note: “Depends where you're at in the state. Detroit is still the automotive hub”

Q How you apply them in the venture world. So that's, that's great to hear. Um, so now you've spent, it sounds like over two decades effectively immersed in venture. How have you seen it change over the years? And has it been for the better or for the worse?

A Well, like anything else, uh, success brings in a lot of competitors, right? We, when we started first round in 2005, we were the only seed stage fund, pretty much. Uh, by 2010, we showed our, uh, investors at the annual meeting a slide that had, here are 12 new seed stage funds starting with the letter F this year. So, so, and, and what happens is, uh, there's a basic law of economic supply and demand, so if the supply of capital Gets too big, right? The prices are going to go, going to get out of, out of range. And so what happened was in, in 2005, I think our average entry pre-money valuation was in the two million range. And a couple of years later it got into three and then it got to five and six and the exits did not continue to increase. And that's the same pattern we're seeing today really, which is that, you know, the entry prices have gone up Way faster than exit prices, which means that returns are going to come down. Um, Josh Koppelman, my co-founder at First Round, uh, an amazing Uh, amazing, both entrepreneur and venture capitalist, uh, had, had a really good podcast a few weeks ago about the venture math for the industry, and it doesn't bode well. Uh, there's too many, there's too many venture funds, uh, you know, and it, it's kind of like when, in 1982, when Jim and I started investing in the PC world, we had 300 companies come to us saying, we're going to have …

AI assessment note: “entry prices have gone up Way faster than exit prices, which means that returns are going to come down”

Q Okay. Uh, as usual for, as part of the podcast, I have so many questions, but let's come back, uh, for a moment if we can to just the, the process by which, uh, the Xylem is helping, uh, businesses draw water. What are some of the challenges, uh, that exists that Xylem helps, uh, markets overcome for purposes of getting access to water?

A Uh, the way I think about it, I break it down into maybe let's call it four categories. Um, you either have too much water. So you're in instances where you're having, um, flooding, uh, and there's too much water going around that you can't manage. Um, that is definitely an impact on businesses. It's an impact on climate. There's disasters, uh, all the time right now that you see with, uh, too much water. We just had, uh, uh, the flood in Texas this last weekend that claimed, um, many lives, uh, as a result of unpredicted, uh, water, um, flooding, 29 feet and upward in a matter of two hours, which Is just unprecedented. Um, so you've got too much water in certain instances. You have too little water in other instances. I lived many years in California and also in Israel, and, uh, you never have enough water, never have enough rainfall. So it's always, how can you get the most mileage out of that last drop and reuse water as much as possible? And then the other couple of areas are, uh, within your water sources, what is contaminated? Uh, and so there's all different kinds of pathogens, um, uh, bacteria, uh, we had, you know, scares of COVID, um, in, in, you know, being passed through, uh, different wastewater streams. So how do you think about making sure that your water is clean and that you can detect all of the different, uh, pathogens or microplastics or pharmaceuticals that…

AI assessment note: “I break it down into maybe let's call it four categories.”

Q talking about, uh, per se is like, okay, well, what about a transmission, uh, you know, energy loss, um, you know, and, and sending in from, um, Electricity from a source, uh, to its user. Um, and you gotta imagine that there's just so much opportunity to, to see improvement. Um, so with that, when you're, when you're looking at companies, what does due diligence look like inside of Nova?

A Well, let me, well, let me start here. Um, I'm representing more corporate venture arm, which is somewhat different to a financial VC. Um, what is the major difference for us? The difference is, and I mentioned that before we are a strategic investor, so we are very strategic in our fields. So the first filter is, is there a fit on the technology side or market side with what we do inside the company? So is that, uh, is there a link to our ceramics business or is there a link to insulation materials? Uh, that is the first filter. If we say yes, there could be a fit We see a combination, uh, or an expansion of our products. Yes. Then we go for it. After that, it's pretty similar to what other investors would do. Uh, we spend significant time, um, on IP. Um, that's usually the, the major asset early stage startup companies have, um, looking into the details of the technology. And then of course, team is always key. You want to Understand the thought process of the founder of the company or the group of founders, whatever applies. Um, and we look of course at business plan. Um, we look at, uh, their customers or sometimes potential customers, um, at their current board and other investors they are working with. I mean, basically the holistic story, and it depends on the stage of the company. Some of them are very early. It's like two people, one patent, no revenues. Uh, that's a v…

AI assessment note: “The first filter is, is there a fit on the technology side or market side”

Q And so for the entrepreneurs who are listening, um, any key lessons or takeaway that you'd share with them, you know, if, if they're interested in engaging, uh, universities out there who have IP they want to bring to market?

A Absolutely. So first of all, you want to, um, understand where it is in the stages of development. Uh, so oftentimes you can have a great idea. Um, and just because it works in a university lab, it does not mean it's commercially viable. Um, and I think that's the thing people will be like, oh, it's validated. It worked in a university lab. And I'm like, that's just cause you can do one sample in a three day process does not mean you can run 10, 10,000 samples in one day. So very, you have to understand that cascade. The other thing is you have to make sure you have the funding to take that idea from a piece of paper to something you can commercialize, and I think that's where a lot of times people miss the gap in how much money they're actually going to need, and it's going to be relatively significant, um, and that's another big one, and then making sure that you're negotiating a contract that you can live to. I think the entrepreneurial mistake is always, I'm going to be commercial in three years, um, and in reality it's going to be five. So making sure that you're giving yourself enough time to take that idea and get it to a place where it's actually generating revenue before you have to begin paying that university back.

AI assessment note: “first of all, you want to, um, understand where it is in the stages”

Q my hat to you for, for being a lifelong entrepreneur. Um, you know, one of the, one of the topics that comes up is, you know, professionals who start later in their career. You know, love to hear, you know, you know, what was it back, back, uh, when you were in your twenties that, that inspired and motivated you to, to get started, uh, with being your own boss?

A I knew I wanted to start my own company. I was raised on a farm in rural Idaho. I didn't label it small business. I didn't understand any of that stuff back then, but my dad was a small business owner. It happened to be a farm, and so I, I knew I wanted to chart my own course. I, I went to college. I skipped a year of high school. I wanted to get away from the farm and, and American poverty as quick as I could. Got expelled from college. Turns out if you, you go to a religious school and you don't plan on following the religion, that, that, that's not a great academic plan. Uh, so then I enrolled at a state school and got a job at an electric sign company to pay my bills, put myself through school, and it was a small business, just the owner and a few employees, and he, he'd been out of the manufacturing game and just did installation and servicing, and after about a year, he, he asked me, he said, hey, you know, what are you gonna do when you're done with college, and I said, I don't know, I just, I know I want to own my own business, and he said, well, hey, if you'll stick around another year, I'll teach you how this business works as long as you'll leave the state, And go do it somewhere else, and I, I was from Idaho. This was, at the time, I was living in Utah, and so we handshook that and moved to Idaho, and being so naive really helped me out. I didn't know how difficult …

AI assessment note: “my dad was a small business owner... I knew I wanted to chart my own course.”

Q And I know your co-founder, who's you on the company with?

A Yeah, yeah. My, uh, my co-founder is, uh, Carlos Santacara. He was a, uh, uh, K-twelve entrepreneur before we got together to build Novel. Uh, he founded a company called NetChemia, uh, that was kind of like an HR software system for the K-twelve space, and ended up selling it to Vista Equity Partners. So, I think we bring to the table as a partnership, or as a, as co-founders, um, you know, I've always been in the, in the funding space, Uh, and he's always been an operator, so we have, we have, you know, two fundamentally different points of view coming at the same common goal, and for that reason, I think we, we've been able to design some, um, some really effective tools for entrepreneurs.

AI assessment note: “My, uh, my co-founder is, uh, Carlos Santacara.”

Q Yeah, I'm thinking about maybe some of our female founders in the audience in anticipation of encountering some of these biases. Any tools or resources that you can think of to help break down these barriers?

A Sure. There are quite a few groups today that are focused on women's entrepreneurship. Uh, there are two that have been around even since I was an entrepreneur. One is called Astia, the other one Springboard Enterprises. Both of those run programs that, that help to train and educate and, and support women's entrepreneurship, but there are many of those today. Um, I, I don't know that I feel like you need to raise your capital necessarily from a firm that invests only in women. I, I know for me, that's not what I was looking for. I think there can be value in that for sure, but just like I believe in the diversity of a startup being beneficial, I think the diversity of your investor team can be beneficial as well.

AI assessment note: “One is called Astia, the other one Springboard Enterprises.”

Q And as a partner, what does your day-to-day look like?

A There are three partners here. There's Mark, who's the founder and managing partner. And then there's Christy who's on the West coast. And then they're both investing partners, but you don't want me picking the companies. Most of my time is spent, I'm more of the finance quant person with a CPA background. So I spend a lot of my time, probably 25% of my time on the diligent side when companies are close, when we're close to closing a deal. And then the vast majority of the rest of the time is just Helping whatever needs to get done at the portfolio level. Since they're nascent new companies, it could be fixing a cap table. It could be setting up a budget. It could be figuring out cash burn. It could be commission plan, bonus plan, whatever it needs to be. If the floors need to be swept, I'll go in with the companies and give them advice to help operate if necessary.

AI assessment note: “probably 25% of my time on the diligent side when companies are close”

Q And love to know, have you started thinking yet about how you intend to monetize and if that's secret sauce right now?

A Pretty open book. Yeah. I'm talking a big game, giving away e-signature for free. I'm giving away invoicing for free. How the heck do you make money? It is on those larger teams. So we have two ways that we monetize. One is collaboration features, specifically between sales, finance, and legal, right? So any, so sales kicks off a contract. Anytime anything's edited or reviewed, or there's a change made, legal wants eyeballs on that. And then anytime it's fully executed and there's an invoice or purchase order created against it, obviously finance needs to be aware of that. So we bring all those together in our own kind of collaborative environment. So we're not competing against Slack, but we have our own collaboration features. And then we charge for that. B to B monthly SAS. And then our ICP, the folks that we're targeting, who we really built the product for is any organization that moves the majority of revenue through a contract. And these are groups to start mid-market, lower mid-market, where they're doing between one and fifty million in annual revenue. That at scale, if you move that much money, you're talking about billions of dollars through the platform, through these contracts, you start to look, act, and walk, and talk a little bit like a bank, but you can monetize in traditional fintech pathways, whether it be The float, uh, whether we think about things like inv…

AI assessment note: “So we have two ways that we monetize. One is collaboration features”

Q view of the life cycle of one of your portfolio companies to say, hey, let me, let me see what I can do to help earn this multiple here. While we're on the subject, I'm curious, what's your sense? Do you foresee it being more traditional advisory, or would you even go so far as to say, hey, we're going to actually See if we can't make a marriage here.

A It's really going to be straight. Sell side advisory, primarily probably 80% of our business, but 20% of our business because of our network with buyers has been either representing or structuring larger transactions. And strategically, it gives us just on the investor side, much better insight to how companies are being valued and sold later downstream, especially with the past couple of years of how The venture market has been somewhat inflated. Big question is where does that right price entry and that entry price. And it just gives us a little bit better insight to where we think we need to be on entry price to ensure that we're getting the kind of multiple that venture LPs expect.

AI assessment note: “It's really going to be straight. Sell side advisory, primarily probably 80% of our business”

Q appreciate that if someone doesn't respond to you, it's not because they don't like you. It's everyone's really busy. Appreciate you're going to need to be persistent and determined. And of course, I think this is inferred, the numbers need to be there. And so on that, what, what counsel would you give to founders just from like growth percentages? What are you looking for? Are there any minimum thresholds?

A There's no official threshold and no one's going to say they have a threshold, but a lot of them do. Pre-seed, it's going to have to be fast. Pre-seed, your base of revenue is small. So if you're at, if you can revenue or something, they may want to see you go to four or 500 next year. It's going to have to be quick. From pre-seed to seed, usually at least two to three or four X year over year. I think from C to A, it'll obviously vary based on industry and what you're spending and all sorts of different things. The general benchmark, I tell everyone below three to five in ARR is you generally need to be doubling every year at a least. I think if you're over five in ARR, it depends on your burn and everything else, but aim to double at least 50, 60, 70% year over year. If you get below that, it can be hard to raise because a VC is going to look at that and be like, how are you going to get over 10 or 20 or 30 in ARR to get me the right exit, even smaller VCs. You have to show that you can actually scale efficiently and get to where they need to be to get the right exit.

AI assessment note: “below three to five in ARR is you generally need to be doubling every year”

Q Got it. Okay. Uh, I want to come back to Inovia and unpack more on the firm. Before we do, let's start with your role. Uh, so you're a partner. What is your, what does the day to day look like?

A That's a good question. Changes all the time. Uh, I'm a partner on a couple of the funds. So on the discovery fund, which is those pre-seed checks plus the fund to fund, uh, side of it. So I spend a bunch of times with, uh, emerging managers that we would invest in across North America. The, for us, that's like fund one through three for them. Usually their funds are sub-fifty million dollars. Um, and we're really looking to sort of anchor their funds and enable them to, to bring in more capital and, and sort of Prop up, not prop up, but invest in the ecosystem in a way like that we want more platform type VC firms around in the same way we were able to do that over the last 20 years. So that's one of the things I spend my time on. The other is on the venture fund. That's really the bread and butter of Inovia, right? So I'm a partner on that. So I spent a ton of time sourcing and talking to founders and sitting on boards and doing all of that fun stuff. Um, I'm also, uh, Part of the CTO office. Cause I spent a decade at, uh, Google overseeing engineering operations with, uh, another partner of mine, Steve Woods here. So I spent a lot of time across the portfolio talking about, um, you know, tech and product strategy. You know, what happens when you were a company that existed prior to 20, 22, like, what do you do going forward? You know, very popular topics of things like that.…

AI assessment note: “sourcing and talking to founders and sitting on boards and doing all of that”

Q Fantastic. So, um, without revealing any trade secrets, um, what's, what's been the key to your success to be able to lower, lower costs?

A So first is that you have to get rid of middlemen, right? You can imagine in a payment ecosystem, everyone that touches the payment makes some money. In your typical payment system, you have someone that sells to the merchant. Who's going to sell card acceptance to the merchant? They take a piece and they set the price. You have a processor that, that takes a price, a piece. A bank takes a piece. Then you have the networks take a piece. So you squeeze everyone out, you squeeze the seller out, the processor out, and in some cases the bank out, and you can deliver an experience where you control every aspect of the cost, and that lets you deliver it. And then when you, when you deliver extreme volume, the networks love you. We do now 70,000,075 million payments a month. This is right in close to nine billion dollars in payments a month. This is new business for the networks. If you look over the past nine years, this is additional money that they are processing through the network they didn't have. This is non-trivial, right? My Visa, I work for Visa, ok? I give Visa so much money, and I thank them every day for the great network they run, that this is incremental business they have because of our services. So as part of it, they, they give us lower rates that we pass on to our merchants. So as you increase volume, increase importance to the networks, Then you can deliver a truly…

AI assessment note: “So first is that you have to get rid of middlemen, right?”

Q Fantastic. What, what inspired you to apply to the YC?

A Oh, well, uh, my alma mater is Airbnb, so I was one of their founding product designers and helped design the user experience alongside the incredible team and founders there. Uh, but Airbnb was a, was a YC company, and I was, uh, we built something really unique, and I was spending time with my, my dear friend, CEO and co-founder of Airbnb, Brian Chesky. I was at his house in San Francisco, and He looked at me after I gave him the demo of our product at the time. He said, man, you're really onto something here. You remind me of, you know, Airbnb right before we did Y Combinator. Like you had a great idea, but you need, you know, YC can really help you accelerate it, tighten up your value prop. Um, and he was right. So we already missed the application deadline as Airbnb missed the application deadline, uh, famous, famous, you know, famous story. Uh, but, uh, there was an exception. Maybe we still had to apply and do all the, the, the, the grind, the application, spent the next few days doing it, and met Gary Tan, the, the president and CEO, went through the interview, and, and fortunately we got in. So, uh, it's kind of a pattern match to, to the Airbnb story.

AI assessment note: “He said, man, you're really onto something here... YC can really help you accelerate it”

Q Who knows what, what the next iteration is going to look like, or if it's going to be considered more valuable than the previous one. So awesome. Um, what, uh, what was the, the trajectory leading up to capital AI? What, what inspired you to start the company?

A Yeah. Well, my background, as mentioned, studied industrial design. I went to a school called RISD, Rhode Island School of Design. Uh, I think it's the oldest design school in America. Uh, I worked on the space program with NASA, designing lunar vehicles and a lunar base, lunar habitat. Uh, and then joined Motorola as an industrial designer, designing rugged mobile devices for companies like FedEx, UPS, the army, and then was recruited to Google and moved out to California and worked on search maps. Um, and then personalization, connecting, uh, Google properties with common identity. I was very fortunate enough to, to work with, directly with Sergey Brin, the co-founder, uh, and then, uh, I built a visual search engine, uh, for comparative shopping for products, art, and fashion, and that, uh, ended up being acqui-hired to Airbnb, uh, where I joined as their founding product designer. Which is amazing journey and building Airbnb was, uh, you know, you know, an experience that still deeply affects me today. Uh, after my time at Airbnb, I, I wanted to design, design a variety of products and really stretch my wings. And so I built a design studio and, uh, that's where I would work with early stage founders, like companies like Brex or modern fertility or product hunt, designing their user experience. Uh, helping them, you know, work, working directly with founders, shaping their …

AI assessment note: “Well, my background, as mentioned, studied industrial design. I went to a school called RISD”

Q Finstrat. Now, uh, that said, to answer your other question about my expectations, they remain super high. Uh, there's, anyone who knows me, this is, there's no, there's no free ride. Uh, very much a capitalist at heart. Uh, someone has to pay, uh, and to your credit, and so everyone's wondering, okay, well, where does this go? So, You're a psych major?

A I am a psych major. I can give the background to make it make a little bit more sense. So I am a psychology major, and I would say I got that gene from you. I think we're a family of deep thinkers at heart. And let's see, it was January that you had offered me the job, but it wasn't until I want to say March that I really thought about it and then came back to you with an answer. Um, and so, the deal was, I would come to Bozeman for the summer, and we would work together, as I'm in San Diego, and the first day that I get there, you're, you tell me, the best way to get briefed on all things finance related, I'm gonna have you, or I want you to, you let me know what you think, you're gonna start studying for the CFA. And I went from psychology to This is actually two months before graduation, because I finished in August, and for seven months, I eat, ate, breathed, slept the CFA, and very smart on your end. I didn't know what I was getting myself into. I truly did not know what I just signed up for, and It was the best way to get fully acclimated in all the basics of finance.

AI assessment note: “I am a psych major. I can give the background to make it make”

Q And, um, I'm going to rely on you here. Everything has to be licensed by the FAA. Uh, and I, is, are, is there airspace that's specific to, uh, this type of work?

A Totally. Yep. Yeah. There's rules and regulations, uh, in every country. Uh, they are slightly different. Canada is very different than the U S. Uh, I am Canadian based, but we're in both countries. Uh, Canada is very much risk based. So we do use micro drones. Microdrones are, um, Such a low risk, because they're below 250 grams, so they're actually treated differently than any other, like any heavier drone. Um, pilots don't actually need a license, although we, we do offer a lot around that space to make sure pilots are, are experienced. Um, Whereas in the U S uh, you do need a license because it's, it's, if you can earn, it's intent based. So if you can earn, uh, you do need a license regardless of the risk. So if you could have a, like a quarter or tiny size drone, but if it's on YouTube and you could make some money, you need a license and that's kind of how it works there. There's hundreds of thousands of drone pilots in the U S that are certified to be able to do that. And, uh, that are on our, you know, 10,000 plus lists of pilots we have right now. Um, And so we're not finding capacity issue at, at all. Uh, and it's actually great because they're trained, they understand, and it's, it's, you know, it's, it's, it's, it's a fantastic system actually, but it, it does require a license. And then, um, the transportation, uh, organization, so FAA in the US, Transport Canada …

AI assessment note: “there is air space restrictions to a certain degree that we abide by.”

Q And okay, so let's, let's take that a little bit further. So was the, the person who won, were they anticipated? Or do you have any sense as to the impact that the technology had on the outcome?

A It was, it was definitely not just like a clear, like they were going to win going into it. And so we definitely felt like we played a big impact in that. Um, we didn't work directly with the campaign, but we worked with a nonprofit group that was, um, you know, Invested in, in getting that justice elected. And, you know, it was really interesting to see how many people actually conversed with the AI agent. Out of the million voters we reached out to, I believe we had 20,000 meaningful conversations. Meaningful meaning, like, four or more messages. And where they asked questions, they asked where they could vote, things about the justice in general, so. Really cool to see that.

AI assessment note: “definitely not just like a clear, like they were going to win going into it”

Q Interesting. I'm, I'm curious, is the consensus that that's a first or second order benefit?

A Um, I would actually say a first because that's something that I don't think any other, um, text provider is doing. Um, in addition to that, Previously, when it comes to opt-outs, I'm sure you've seen it when you get a text, it says, you know, reply stop to stop to end or stop to opt-out. If you don't text back stop exactly, it doesn't count. It doesn't register that that's an opt-out. Whereas our system, you would think it's actually kind of shooting us in the foot and counterintuitive, but we opt-out other things. If you text back S-T-I-P, because you misspelled stop, We'll still catch that. Or if you say something like stop texting me. That's something that AI is able to interpret exactly as we need it to.

AI assessment note: “I would actually say a first because that's something that I don't think”

Q I'm curious, do you remember what, what it was that sparked the interest in working for Dolores?

A It was, I think, like, many middle schoolers. I think it was a combination of Law & Order SVU and Legally Blonde. I think I saw it on TV, and I was like, that's very cool, but I am appreciative of my parents that they encouraged me and, like, drove me to the DA's office, and I'm, like, forever grateful for Dolores for, for letting me participate in the, in the DA's office throughout my childhood. It's, like, truly what I did. It's where I would go after school and It's, if there was a hearing that I wanted to continue to listen for, like, an entire course, uh, cases, the course, I would miss school, and I would go watch it. Um, I would, like, I was, I learned how to drive with the DA's office attorneys. It's, like, truly where I grew up, and that's a very unique opportunity that I had, and I'm forever grateful for the Douglas County DA's office for, for letting me do that. It changed my life.

AI assessment note: “I think it was a combination of Law & Order SVU and Legally Blonde.”

Q Agree hundred percent. All right, so let's talk about, um, the lead in to Soxton. Um, you mentioned that you started some companies live in, in Boston. What did that, what, what were the companies, what was the experience like, and how did that inform how you run the company today?

A Oh my goodness. Oh, every bit of it. So I'm very proud of like the first company, which still exists that I started while in law school. It's called Spencer Jane. It's a machine washable pantsuit. And so when I was in my first year at law school, I needed a pantsuit and I was like, this is an incredibly frustrating experience. Like, where does one find this? Why are they all ill fitting? What's the, what, where do my friends go? Apparently this is a common problem. And I really just, like, couldn't get my mind off of it, and I started across and rolling at HBS, and I ultimately created a, a pantsuit that is machine washable. The arms are intentionally longer than other, than other sleeve links because folks don't want to get them tailored, and so they're designed to be rolled. They create an hourglass figure, like, it's like a whole thing, but I learned so much in that process. I learned how to, like, source fabrics. I, I found a fabric that I liked from a fancy brand. I, Added everybody on LinkedIn from this brand until somebody told me the fabric mill, which is placed in Italy. And they connected me with a factory who had a son my age who spoke English, and then the parents, like, were, like, so nice, they're, like, okay, well, like, take a chance on you, and so my factory is the same factory that does, like, Dolce & Gabbana, Montclair, and all these fancy brands, and, um, so…

AI assessment note: “I'm very proud of like the first company... called Spencer Jane. It's a machine washable pantsuit.”

Q Okay, I'm gonna let you off the hook, and I'm not gonna ask you for any, any cons, just pros, um, but you're welcome to share your cons if you want, but I presume you're working with all the major models?

A Uh, yeah, yeah. So, uh, what we're doing is we are, We work with any model, so we're model agnostic. What we do is our own models, they are small language models. We're not using large language models. So yeah, so what we've built is a suite of, um, 17 small language models that act as a sort of guardrail on top of the LLM. So what, what usually happens is, uh, we install our small language models on top of the Um, on top of AI agents or on top of the chatbots that employees are using. And what ends up happening is, uh, the, for example, the, the, the admins or the managers can control what kind of, uh, guardrails they want to, to use. For example, we did code detection. So for example, if someone sends a piece of code to the AI, we can detect whether that's a sensitive code or it's just like random boilerplate code. That's one thing. We also mentioned the anonymization. Uh, so that happens. We can detect like, um, um, An emotional language or, uh, a prompt in a different language. So there are different, uh, guardrails that we have, and those guardrails are based on small language models. Um, and why we're using small language models, uh, I think it's very obvious. They're explainable, they are, uh, efficient, uh, and, uh, they're auditable. So our models, for example, we, the, the minimum requirement for us to run a model is, to run a small language model is basically, um, a,…

AI assessment note: “We work with any model, so we're model agnostic.”