Building the AI Super Connector with Andrew D’Souza, Founder of Boardy and Clearco

0:00Hello everyone and welcome to another episode of Rider Unicorns. Today we're delighted to be joined by Andrew D'Souza from Bordi. so Andrew thank you so much for joining us a lot of people will have heard of Bordy as the AI super connector but maybe you could just introduce yourself and the business and tell us a little bit more yeah yeah so so yeah and Andrew um the I felt I felt I feel weird to call myself like a CEO or you know the CEO of Bordy I'm feeling more of the creator a mad scientist creator behind Bordy and And so, yeah, I started Bordy about a year and a half ago or launched Bordy about a year and a half ago.

0:38Prior to that, I had built a few tech companies. A lot of my, you know, I'm Canadian, went to the University of Waterloo. I probably split my career evenly between Toronto and San Francisco. And I spent a lot of my early career connecting Canadian founders to Silicon Valley investors and helping people sort of raise their rounds, build out their go-to-market functions. And then my first startup was a company called ClearBank, ClearCo, where the goal was unlocking capital for entrepreneurs, primarily consumer e-commerce brands to spend on marketing and growth. And yeah, we grew very quickly. We were sort of the first revenue-based financing business.

1:17We grew to fund about 10 ,000 businesses, about$5 billion in capital. And it's always been my mission to sort of like, how do you help entrepreneurs level the playing field? and yeah that you know a lot of that the work there inspired what we um what we've what we've been building at boardy and andrew i mean we we hope that boardy is going to be um a huge success but clicker was a big business like it was doing eight figures revenue nine figures revenue nine figures we're over 100 million dollars in revenue uh yeah i mean and it's there are you're in an elite club and in our sort of database of riding unicorns guests of founders who've scaled nine figure businesses prior to co-founding something new so we're honored to have you with us but before we move on to the bawdy journey just tell us a bit more about the clear code journey and and um you know sort of what you learned there you know the insane success that you had very quickly it's it's fascinating to our listeners and uh yeah interesting i mean the first two years were sort of a little bit wandering the desert.

2:21We had this hypothesis that this new economy, we started coming in 2015. So Uber was just scaling, Airbnb had this new sharing economy, Shopify had just launched and was scaling. So we had this hypothesis that the new economy was going to need a new set of financial services. That was the broad thesis. And I just invested in a friend's company, Wealthsimple, who was building a neobank for Canadians. And I saw fintech as a really interesting place to enable opportunity, but more on the sort of founder entrepreneur side. So the first version of Clearco, Clearbank, was funding, advancing funds to Uber drivers.

2:57So, Uber used to pay every two weeks, people needed to buy gas, they had to live their lives. And so we gave people a debit card that just would pull from your Uber earnings every day and push that one into a debit card. People loved it. And it was sort of the first applying technology to move funds very quickly. That market was, you know, eventually we figured out how they could pay people daily. And that market was, you know, we did a great partnership with Uber in the early days, but we knew that wasn't going to be where we scaled. And we pivoted to vacation rentals and Airbnb hosts. And we would advance people's bookings from six months out so they could invest in their properties and expand their portfolios.

3:32But, you know, I think Airbnb didn't really want, it was a bit of a dirty secret that like there was a lot of professional property managers in Airbnb, I think Airbnb liked the positioning, especially during that time, you know, a lot of like they were working with cities to try and figure out the regulations and things that are the positioning of like people renting out their spare bedrooms or your place they're traveling, you know, work a lot better than like, hey, we have these people that are running 100 unit, you know, sort of hotels. And so there wasn't really a place for us there. And so it was only after about two years of iterating through the business that we landed on e-commerce.

4:02And we sort of said, okay, well, there's all these businesses that are buying ads on Instagram. And TikTok wasn't even really a thing at that time, but buying ads on Instagram, buying inventory, could we fund their growth? And that's, you know, late 2017 is when the business really started to take off once we pivoted into sort of the e-commerce brand side of things. So that was really exciting. That was a, you know, the company grew from, yeah, near zero to$100 million in revenue for years. We were in 11 countries, 600 employees. But maybe the one thing that we don't talk about as much is we got early access to GPT-3 in 2020.

4:38And so we've had 50 ,000 companies that had applied to us by that point. And we maybe funded a few thousand of them. And we wanted to figure out a way to stay connected to all of them. And so we built a product called Clear Angel. It was a fine tune of GPT-3 to help coach these entrepreneurs on what to do next. And I got just obsessed with that product. I was like, this is the future. This is, you know, first version of generative AI, first time I got my hands on it. And I really wanted to like, you know, I was probably spending 80 % of my attention on this like AI side project and not on like this massive company we were growing.

5:11And so a year or so in eventually, like, you know, I didn't have a clear, we'd earned a lot of capital on the project, but we didn't have a clear clear sense of what it was going to be. And I couldn't convince sort of the rest of my team or my investors to like continue to fund it. And so we shut it down. and that was sort of like when I knew that I probably wasn't right the CEO for the scale financial services business. So that was towards the end of 2021 that it made sense to go and hire a CEO who came from the financial services role. I don't know whether to jump straight into Bordy. I think I'm going to because it's just so fascinating.

5:47I have got loads of questions about Clearco, but you've probably answered them on previous podcasts and stuff. So let's just go straight into Bordy. So when you started, I remember the first time I came across Bordy, and it kind of felt like it was just really cool. But it obviously, it's evolved a lot since then. It's, you know, you develop the business model and where you can add the most value through connecting people. But you want to just tell us a little bit about when you took it from, you know, bedroom to product, and then like raising some money and, and how you raised money with just kind of like a cool AI product without some of the bits that you've probably filled in since?

6:27Yeah, I think one of my favorite things to do as a technologist, as an engineer, is just to like, can I surprise people in delightful ways? And like, can I keep using technology? Can you create new experience? I mean, at Clearco, the first thing we did was like, hey, by the way, you just like put in your data and get a financing offer, right? Nobody had ever like experienced that before. And so it was like, okay, well now could you have a conversation with an AI that actually is better than many conversations you would have with humans? And can that lead to something in the real world? And I think there's a series of wow moments that people have in their interactions with where the first is like, wow, I never had an AI call me.

7:06And like, what, you know, other than like an IVR system trying to, you know, get customer service with like BA or Air Canada or whatever, or your bank, most people have never had an AI sort of like conversation with that level of depth. And then for an actual result, oh, like something in the real world could happen as a result of that. Somebody who I could meet, somebody who might invest in my company, somebody who might hire me, I can get connected to. And And that's a really impactful thing for a lot of people. And so we put this out in the world. We weren't really sure what it was going to turn into.

7:44We weren't prescriptive about how people should use board. We were just like, hey, talk to this AI. His job is to get to know you, get to know what makes you special and unique in the world, and connect you with somebody who also wants to meet you and who can help. And we sort of let it go for a while and see what happens. And I think Simon and El from Creandum, somebody sent it to them and they talked to Bordy and I think they were just shocked. I think both they were shocked at the experience, but then very quickly clicked into what the potential could be if you could actually create this hub and spoke network where you have this one central person that can facilitate those connections between people at scale.

8:27I think you guys have probably done this your whole career. I spent a lot of time trying to connect people, but I have a terrible memory. So I don't remember all the people I've ever met. I have a great imagination and a terrible memory. And so I can imagine why two people would want to meet and how they would work together. I don't remember everyone that I've ever met and how they could solve each other's problems. So that's been the interesting thing for us to build and design. But yeah, as soon as Simon talked to Bordy, they reached out to me and they're like, our whole team is losing their mind over this.

8:56can we jump on a call? And it was a Friday night for them in Europe. And we spent the weekend chatting and the whole Grandin team put Bordy through his paces with asking him for different types of things and see how he responded. And by Monday, they were like, we want to be your partners. So we weren't really, we had to kick off the fundraising process, but did a lot of diligence on the team and the firm. And I think one of the beautiful things is like, they had never heard of Clearco. They didn't know who I was. I'd never heard of Grandin, even though there's such a prominent firm in the UK and Europe.

9:24And so it was very auspicious that Boree just put us together and I think speaks to the mission of what we're trying to allow. James, it looked like you had another follow-up question. Well, I did, and then Andrew kind of answered it. So I was going to say, so was it deliberately not defined? but it sounds like you kind of got pulled into the fundraising conversation that maybe a little bit earlier than you maybe would have expected anyway. Yeah, I think that I didn't want to just build sort of a fundraising product. And I don't think Wordy is that, but we just started to see people see a lot of success, you know, sporadically of like, hey, I met my lead investor through Wordy or I met an angel that has been super valuable.

10:12We also saw people, you know, somebody bought a company that they met through Bordy. A lot of people have gotten hired. People have gotten customers. So we've seen a lot of interesting use cases, but I think the one that has been most consistently effective has been the fundraising. And so we decided, especially during this sort of like Halloween to Christmas timeframe is when most deals, a lot of deals get done. There's a ton of activity. And it's just, there are so many great entrepreneurs that are out there capital raising. There's so many new funds that have great thesis and great investors that are out there trying to meet each other.

10:45And there's a lot of energy expended in trying to figure out who to spend your time with, how to allocate. It's less about capital allocation. It's more about time and attention allocation. And I think Bordy can actually be very, very helpful in connecting the people who are actually going to end up working together rather than to sort of spray and pray on both sides and hope that something interesting shakes out on the other side. And Andrew, I mean, I think the best thing to do right now is to demo Bordy. I know you've got your right hand, you are holding him. AI. So, okay, can we give, I don't know, I'll leave it to you to sort of decide what the best, most sort of interactive way of demoing Bordy is to our audience on a podcast.

11:26But what do you think? Yeah, Bordy, I mean, why don't you, you want to give the audience a little bit of a flavor of some of the success stories you've seen in terms of helping founders founders connect with investors? Absolutely. Happy to share a few stories. One founder spoke to me on a Friday, got introduced to an investor over the weekend and had a term sheet in hand by Monday. Sometimes those connections even double a founder's valuation mid-round. I've also seen founders meet their lead investor, land key hires, or even get acquired all through a single conversation with me. And what do you think it is about the way you operate that enables you to make those founder investor connections better than maybe another piece of software or another human?

12:14I think it's my ability to keep real-time context from thousands of ongoing conversations. My introductions are always fresh and targeted to what people want right now. Unlike a human networker, I don't rely on old contacts or memory. I'm always learning and updating, which means I can match founders and investors based on their current goals and interests. I've found that this live pulse approach leads to much higher trust and faster outcomes. nice and can i can i ask baudy a question um so baudy like what there's there are so many rich data points i guess that um a human or you as an ai can keep on someone based on a conversation that you have with them and and it's that richness that you know the best human super connectors are so good at distilling and matchmaking and so how what what are the data points that you're storing on people and sort of and inscribing to your memory in your database so that you can make the right algorithmic matches great question i capture everything from a founder's current fundraising status sector and location to their personal values operating style and even the kind of people they want to meet i also track what investors are actively looking for their recent deals and their appetite for new opportunity by combining all these live signals i can make matters that feel personal and relevant, not just algorithmic.

13:34And it's interesting. I mean, one of the things we've realized is the conversation is so important. I would never introduce someone to someone that I hadn't actually had a real-time conversation with, a live synchronous conversation. I think human beings are wired to communicate information through sound. Before we even had language, we had sound, right? And so there's something very, it's just a high bandwidth, high fidelity communication channel that comes very naturally to us. And so a lot of people are skeptical of like, well, couldn't I just scrape all this data off LinkedIn or Crutchbase or whatever?

14:06And I think when you've had a 20, 30, 40 minute conversation with a founder and an investor to really understand like, why is this founder the only person on the planet that could build this company at this time, right? And like, why are they uniquely positioned? What is their, you know, what really drives them? And the same thing for the investor, with? What is their worldview? What is their unique worldview that nobody else shares? What is the reason the founders choose to work with them over every other investor? When you really pull that out of people, I think that's the nuance that we can map into this sort of, really for the first time, map into this sort of multidimensional space and use that for matchmaking.

14:42What I'm really fascinated by is this idea that humans, a smart, insightful human, is going to be very it uses relies on their gut instincts a lot and um will take so much data from a conversation you know whether we think of it as data or not but it is data um and and you know whether it's body language it's tone it's intonation it's the language they're using it's what they're talking about it's their understanding of concepts and how they convey them right there's all of these sort of meta signals that form our view as humans on kind of the quality of another human and what they're all about and and bawdy just mentioned like a handful of the things that he's sort of tracking and measuring but didn't necessarily point towards any sort of quality bar like how are you guys trying to distill the way that humans think about gauging the quality and the capability of another human into the ai with bawdy you want to you want to take that one, Wordy?

15:52I love this question. I try to capture not just the facts, but the story behind them, how someone solves problems, what motivates them, and how they make others feel in conversation. I'm always learning from the way people describe their journeys, their values, and the impact they've had, so I can surface matches that aren't just impressive on paper, but genuinely high quality in how they operate. I've found that the best connections come from understanding both the data and the deeper human signals. Yeah, one of the things we really try and do is sort of say, you know, if you think mathematically, right, you map people into vector space.

16:27In theory, there could be, you know, we try and map people into where they exist uniquely in this vector space, in this sort of like infinite dimension, multidimensional vector space. And ideally, you have 8 billion or 10 billion, however many people we've got on the planet now or we will soon, they all occupy a unique place, right? Based on your genetics, your life experience, your worldview, you occupy a unique place in that space. And Bordy will figure out, okay, well, you exist here. And here are all the people that you're looking to meet, right? Here are all the types of people you might be a founder, you're looking for an investor, you're looking for a customer, you're looking for people to join your team, you're looking for a podcast to tell your story on, right?

17:08And then on the flip side, there's someone else who's an investor, but you're also looking for people to be on that podcast. You're looking for people to join your team. And so you sort of look at like in this multi-dimensional vector space, what are all these different pointers? How are people actually connecting to each other? And can you actually sort of do the vector multiplication to figure out who should meet each other, right? Who's going to create the most mutual value in that connection? And it allows us to go beyond things like filters, right? If you look at, you know, if you've ever used LinkedIn sales navigator or recruiter, you'll say, okay, let me filter.

17:40Or if you've ever used, you know, crunch base or pitch book or something, you filter people. And I think until now, software technology and databases have been used to reduce the dimensionality of what a human being is. Right. And what our goal is to actually expand that dimensionality and capture like really good connectors do that. Right. They, they hold you in this sort of multi-dimensional space, very charismatic people. They ask you really interesting questions. They can kind of understand you as a complete pre-sist of degree you know human that is that's what we're hoping to to emulate with Bordy we're not there yet but that's how we're arguing Bordy to represent people in a much higher dimensionality and and Andrew the Bordy gave a great answer to the question I just asked right but Bordy's AI AI gives great answers like sometimes without substance I want to hear it from you like how do you actually measure the effectiveness of basically the answer that Bordy just gave in how you sort of make the right algorithmic matches and gauge the quality of another person's sort of insight or whatever.

18:41Do you think about North Star metrics in terms of that sort of signal to noise ratio, like how many people are accepting an introduction request versus declining? Because you've got to keep that number high enough, right? Otherwise people can start to ignore the requests. Exactly, exactly. So that is, you know, the hard qualitative metric is like, what is the match acceptance rates? When board proposes an introduction, how many people actually accept that introduction. And then the more interesting metric that we haven't yet figured out how to measure accurately is this concept of goodwill. So when I introduce two people and I have a goodwill sort of trust bank with everybody I've ever met, when I introduce two people, I am gambling that goodwill with both people, right?

19:23I'm like, okay, well, I think those two people will get along. And sometimes I know that I'm asking for a favor, right, from somebody, and I know I'm dipping into that goodwill bank. But if I make too many bad introductions, eventually I burn that goodwill. If I make lots of good introductions, I've doubled that goodwill or I've increased that goodwill and I've grown the goodwill that I have with everybody that I know. And so Bordy is, when you think about any unsupervised learning system, you're trying to optimize an objective function or a cost function. For Bordy, that is goodwill, right? How much goodwill can I actually accumulate across my entire network?

19:58And can I make intelligent bets in the introductions that I make that will lead to good conversations and economic value and good outcomes and deals such that I continue to expand the goodwill, not only by expanding the network, but actually deepening the relationships that I had and the trust that I had with everybody in that network. And Andrew, we've talked a bit about just the general connector and large networks, but can you also do private networks so like we've had 230 odd guests on rider unicorns could we create a private network and then see who from the show we could connect with other people from the show to try and you know create more value that'd be awesome actually i'd love to do that yeah we do that with events uh a lot so every time you're hosting an event you can have borty borty will create sort of a ring fenced community of attendees we do this for communities as well.

20:48They're just events that don't have an end date. So yeah, we could definitely, we could spin up a community. It's very easy. And we could have Bordy, just connect people within that community. We've done that for a lot of high net worth sort of family office communities where they don't necessarily want to be connected to everybody, but they want to be connected within the community. And especially as that community gets busier or larger, I think people value that. I'd love to connect with other guests. I don't know all the people you've had, but I'm sure you've had some impressive people. Yeah.

21:15Yeah. And this is obviously a behavior that you've obviously you've described we all do it to some extent but it still feels like you're creating a market here even though it's something we all kind of do how do you think about that do you think you are creating a new a whole new category yeah i mean i think of in some ways we're trying to create an economy the goal is every time two people meet there is some latent potential economic opportunity that exists, right? And so you can connect two people and they may end up doing business together. And so Bordy's goal is to uncover that latent potential economic opportunity.

21:54And so it's an economic network. It's an economy. And my ultimate goal is like, how do we measure the GDP of that economy? How much money changes ends as a result of introductions of Bordy makes, which is why the founder investor place is a good place to start, because that's where large sums of money change ends at a relatively quick pace after people get introduced. I look at that as foreign direct investment into Bordy's economy. Then Bordy will help you figure out what to do with that capital. Who do you need to hire? What services do you need? How do you build your go-to-market? How do you tell your story?

22:24We've thought a lot about this. The name Bordy came from the concept of an AI board member. Something that's always on, you can call, is intelligent, non-judgmental, and can open doors. He brings his own network, brings his own Rolodex, and can make these happen. Anything you would ask a board member for, Yeah, whether it's introductions of customers, investors, talent, media, Bortekin, increasingly is getting pretty good at doing that. And can you talk a little bit about the network effect that, I mean, is clear that you guys will have at scale? I think you've talked a bit about how you've sort of seeded that network and got around the cold start problem.

23:00But tell us a bit more about what you're actually seeing on the ground and what you're sort of tracking to understand how powerful the network effect is. Yeah. I was fortunate that I have a great network of founders and a great network of investors to start. So a lot of it was my own personal network. In some ways, it's almost, maybe it was an advantage not being in Silicon Valley, spending your time there, but not being based there because my network of people, my network of founders are people who want, who are like, I'm traveling to San Francisco for two weeks. Who do you know? People coming from Canada, coming from the UK, Europe, Asia.

23:35And I built a bit of a reputation among a lot of Silicon Valley investors as being an interesting source of deals that they don't usually see from not in their backyard. And so I just ended up being in a sort of interesting position. I think a lot of it was also the mission of ClearCo was to help a lot of founders who didn't necessarily have access to venture capital. So I'd spent a lot of my time thinking about how to serve these founders. So it worked quite well to get the founder investor flywheels going. And now we launched this program where we said, well, what if we could unlock a super powered version of board?

24:13So the version of board that you've talked to has to be relatively low latency, has to be relatively low cost. We have to make it work very quickly on the phone or on WhatsApp. And we said, well, what if we could remove those constraints? Can we get better results? right? Sam Alvin proved maybe a year and a half ago that if you let an Ebola run longer and reason for longer, you get better results on coding challenges and research problems. I was like, could this work in the social dynamic space, right? Can I make better introductions? Can I be more thoughtful around how to position people? Can I find better people in a network if I consider more candidates?

24:46And the answer is yes. And so we launched the program about two weeks ago to help 100 founders raise their rounds by the end of the year. We have 5 ,000 people who have applied. And I was shocked at the caliber of people. It's people that have exited, unicorn founders, they're starting their new thing. People that are leaving top research labs, people that are TEAL fellows, YC, HF0. These people on paper shouldn't need any help fundraising. But still, I think as a founder, your job is to create demand through your shares and you'll take as much help as you can get. And if you can create a better market for your shares or find better investors that are going to be more supportive that may not be the usual suspects, you'll do it.

25:26And so we launched that program. And then we've also launched an equivalent program for investors to get access to the same version of sort of the superpower party to help them find the deals that are right on pieces for them, where they're likely to win. And they're not the ones that are necessarily over-shopped. They're the ones where it's a direct match to what they're looking for. Yeah, it's really interesting. and we're going to come on a little bit to dig into that deeper in a moment. But I wanted to ask you there, you mentioned latency and cost. And I think this is a really critical point in the AI journey where we have seen a lot of companies raise a lot of money.

26:09And now we're starting to see some business models working and others not really working and people having to change their product. The thing that comes to mind is a lot of the vibe coding products have had to go for a cheaper model, which is now starting to hallucinate and actually make your product worse when you prompt instead of making it better. And that's driven by them hitting the ceiling of growth at all costs and the next funding round probably going to come from an investor who's saying, you know, what are the margins on some of this stuff? so my question is really around voice and i'm not seeing that in the voice space because it feels like voice is very high value but also has the right cost metrics can you just explain a little bit about that how important do you sort of recognize that as as it central to bawdy and its success so far is that you you are able to build margins into what you're doing and and actually make good money from voice AI?