Episode Summary:
In this episode, Eric Weiss sits down with Andy Ballester, serial entrepreneur and co-founder of GoFundMe and EyePop.ai, to unpack the highs, lows, and lessons of building category-defining startups. Andy shares how an early college job at a software startup sparked his passion for technology, why he bootstrapped GoFundMe into a global fundraising platform, and what it took to scale from two founders to a household brand that processed over a billion dollars in giving.
Andy also reveals the insights behind his newest venture, EyePop.ai, a computer vision platform making AI accessible to developers and businesses across industries. He explains how modern visual intelligence can turn everyday images and video into actionable data—from analyzing pickleball swings to inspecting drone footage of rooftops—and why speed, lean teams, and clear product design are essential in today’s fast-moving AI landscape.
Listeners will learn:
Whether you’re a first-time founder, a product leader, or an investor tracking the next wave of AI infrastructure, Andy’s journey offers a masterclass in startup strategy, growth, and perseverance.
Eric Weiss (00:01.916) All right, welcome. am here today with Andy Ballester, a serial entrepreneur here in San Diego, co-founder of GoFundMe and iPop AI. Andy, thanks for being on the show today.
Andy (00:13.506)
Hey, thanks for having me.
Eric Weiss (00:15.588)
Awesome to have you. you know, you've been obviously kind of one of the, the, big names in town for quite a while, but sort of building in the background. you know, go fund me, obviously you were in the lab, you were growing this, this huge business. now you're, you know, starting the new thing in AI and I'm just really sort of excited to hear what is your entrepreneurial journey? Like what got you started even before go fund me, but what, kind of lit that fire for you?
Andy (00:40.94)
Yeah, the actually the first startup I did was in in college. It was a, know, I had to had to have some sort of job in college. And, you I don't even know if Starbucks was around in in the Midwest, in Illinois at the time I was going to college. But, you know, as opposed to going and getting a job at like a Starbucks or coffee shop, I was able to find a a job at a startup in software development and
that really kickstarted the whole thing. I was able to see the intersection of taking technology and making it accessible to users who were new and at the time in early 2000s, they were new to the web in general. And that journey really, really started me off down the path of I want to do startups. It's amazing to see the ability to take a problem.
Eric Weiss (01:25.16)
you
Andy (01:38.338)
to code all night on it, and then the next day have something in user's hands with those early web technologies. It solved a distribution problem, and you could really ship code very quickly and get it in the hands of folks and get that feedback really rapidly.
Eric Weiss (01:56.304)
Yeah. Yeah. I followed a pretty similar journey. mean, I was working at like internet service providers, slinging racks of servers and network cables and stuff back in the nineties and early two thousands. And, I definitely always had that kind of that, I don't know if it's impatience or that, that need to just do things quickly and get things done. And so that kind of brought me in the same into the startup scene as well, again, well over 20 years ago. but so then, so then
Andy (02:04.59)
Yeah.
Eric Weiss (02:22.802)
Tell me the journey of, what made you decide to then go from that to starting your own company.
Andy (02:28.462)
Yes. So I moved after college, I moved to San Diego. I was able to find a job at a startup here and I had the really unique experience of seeing both hyper growth and then what it looks like when in the opposite of the hyper growth scenario. And so, you I was employee 35 at that company. We scaled very quickly to 300 employees and 50 million users worldwide.
And then because it was a mobile-based company, just with the speed of innovation in mobile, didn't make the turns fast enough at that particular company. so what happens when we don't have the revenues and we have to lay off our teams and roll that back down to 50 employees if that kind of tells the story of where we were. All right.
Eric Weiss (03:21.456)
Yeah, was this around 2008 or?
Andy (03:24.814)
That was exactly around 2008. so that's, yeah, it's a rough time for sure. So going through that journey, really cemented the idea that you can build just incredible things really quickly using software and the web and all of these new practices. And you can also get your product in the hands of everyday folks really quickly. And so I knew
Eric Weiss (03:28.444)
Fun time for all of us.
Andy (03:53.998)
I kept, I wanted to keep doing that. And so in 2008, when we really didn't have anywhere to go with that startup, I talked to Brad Damphouse, my co-founder for GoFundMe, and we took a stock of where we're at. His startup also imploded at the same time. He had moved and gone up to LA to do a couple startups, and then it just happened to be...
not fortuitous timing, I guess serendipitous for us that we got to work together again, even though, you know, it was around kind of two startup disasters. And we said, well, you know, we, we've put all of our energy into these, these previous startups. We were both pretty broke at that point because we've been getting paid in IOUs. And so I hope we're like a natural conclusion. Let's start another company. Right. And so let's start something together. Let's use the lessons we've learned from.
being in the startup world and let's see if we can do this ourselves.
Eric Weiss (04:57.649)
So what was the opportunity that you found that was, hold on a second, let me pause. I'll edit this out.
Andy (05:03.006)
Sure. Yeah, totally. No worries.
Eric Weiss (05:16.313)
occupational hazard of having a doc. All right, let me let me back up. So, so what was the what was the opportunity that you saw that started GoFundMe?
Andy (05:17.758)
Yeah, totally. Sure.
Andy (05:27.83)
One of the things that had gotten Brad and I through the ups and downs of startup life and yeah, sometimes you get paid, sometimes you don't get paid. You don't totally know where that next paycheck is coming from is we were big savers. were very frugal East coast folks, you know? And so we were looking around and we saw our friends putting a lot of things on credit cards, having high car payments and
and just juxtaposing that with what we have been doing in our financial lives. And we were big believers in setting goals, like personal goals. And so we thought, you know, that's interesting. like there's a mission there around helping people save money. One of the things we also learned is if we were going to start a company at the time, it was very much in vogue to start a company where, you know, just get a bunch of eyeballs.
question mark, and then we'll make money later. And so we didn't want to do that. We saw the power of being inside of the transaction for FinTech. we had that learning sooner than a lot of other folks, because we saw it in the mobile space. And people don't think about, especially early mobile, as a way to transact. But there was a way to transact on premium text messages.
And so we saw the power of being involved in that transaction. And we wanted to get into the burgeoning fintech space of where everything was going. So we tried to figure out, OK, what's the unlock here? What does that product look like? And our first product was called CoinPiggie. And it was a way for people to save. I think, here, I'll just describe it really briefly. And I think you'll see kind of the origin here.
But say you wanted to save up for a TV. That's great. We want you to have that TV. But let's figure out a way to save up for it instead of just throwing it on a credit card. And so you would put your goal into our website. It was very gamified. You would have a page that was both your, it served as a dashboard, as also a public announcement of what you're saving up for. And you could share it or not.
Andy (07:48.27)
And then you would save on a monthly basis if there was ad hoc things like You happen to to get a little money in there. You could you could push it in on an ad hoc basis But there was a monthly savings recurring thing that you Basically take money from your bank account and put it in there and that would reflect on a meter on your page Towards your goal now in the background we tapped into other API's and were you know?
going to try to get you a deal on that TV so we could reduce the goal potentially that you had to save up for. And so there's a lot of different things we wanted to pull into there. And there was this one use case. And this was really important to us. One of the other components to a product that we were going to build was having to do with the viral loop. We had built, Brad and I, insane viral loops in the past. That's one of the reasons our previous startup there, we worked together.
got so big. So we wanted to pull that in. So we were thinking, does this become social? And so if you're saving up for that TV and your aunt wants to kick in $20, you could send this page out to your friends and family, maybe for your birthday, and they would kick in $20. And so what we learned in building out that early proof of concept there, both on paper and then in a POC in code, is that
fintech at the time, especially for two people who are not venture backed, who have no affiliation with banks, it was going to be really tough to make a good savings account there. In fact, because you were dinged on the money that you're putting in for interchange fees and all kinds of other fees, it was almost like a negative 3 % interest on the way in. And so that was not going to fly. We didn't think users would really like that. But we did have this use case of, well,
Eric Weiss (09:34.373)
Wow, yeah.
Andy (09:42.124)
what happens when somebody else contributes money to the thing you're saving up for. I don't really care about the small ding of fees there because it's kind of found money. And so we thought about that more as what if that was the primary use case? What would people collectively save up for together? And that's how we thought about this concept of personal fundraising in the early days is sort of collective saving.
Eric Weiss (10:10.15)
Yeah.
Andy (10:10.184)
And then when we actually built out the product around that concept, we realized it's a form of fundraising. so before, certainly, we knew that the term crowdfunding, that we could get coined later and we'd be lumped in with the crowdfunding crew, we thought about it as personal fundraising. What would you personally fundraise out there for in the world? so we retooled CoinPiggie.
Eric Weiss (10:17.968)
Mm-hmm.
Andy (10:39.766)
scrapped that, started a new project called Create a Fund. And we brought over a lot of the early elements that we thought about. And so you had a meter, you had something you were collectively fundraising for. We had the ability to tap into your networks. And one of the networks we were really excited about, and this is 2008, remember, is a new one, you know, the new one, which was Facebook. Not a lot of folks had
Eric Weiss (11:03.089)
Mm-hmm.
Andy (11:08.352)
started tapping into the graph at that time for this type of activity. And that was something that was really exciting to us because we had built our viral loops in the past off of email. And this offered a new modality in which to spread the word about what you're saving up for. And we thought early on that people would use this to help themselves and better their situations, to help their friends and family.
and to help their community with different causes. And that early idea of Creative Fund actually is very much like what GoFundMe is today. We pivoted a couple of times away from that idea and back and forth, but that's where we started.
Eric Weiss (11:57.224)
Oh, that's what I mean. Yeah, they always say, you know, the entrepreneurial journey, it always zigs and zags. But yeah, amazing. Sorry to do this. Can you give me one minute? I'm there's somebody at my door and they're that's why my dog is going crazy. So I will be right back. I'm sorry.
Andy (12:02.731)
Yeah.
Andy (12:12.142)
Yeah, yeah, no worries. No worries. Yeah. Okay, hold on and just grab a water real quick too.
Andy (12:56.59)
Thanks
Eric Weiss (13:43.73)
Sorry, a worker who was supposed to be here later showed up early and then is like, where are you? right, don't worry. I can always edit this stuff out. That's a nice thing. We'll select and cut. All right, cool. we've got the origin story of GoFundMe. So obviously, that was a huge kind of growth story. What were some of the biggest challenges or growing pains that you faced?
Andy (13:55.295)
Yeah, no.
Andy (14:11.51)
Yeah, so initially the biggest challenge was go to market and not just the traditional, we have to get the word out, but also there was a lot of technology issues we had to solve in that time frame. And so when you look at our first version of what this concept was in Creative Fund, that
That was plagued by just early fintech features, and they were not ready for us. And so we had to figure out, how do we move money back and forth without trying to aggregate it ourselves, which is illegal unless you have a whole bunch of bonds and things. And so we had to tap into payment providers at the time. And so PayPal was the best game in town, but they really weren't ready for that early use case.
Their fraud models were all built around people transacting on the internet, primarily around the eBay use case. And when we started getting the early small traction, we faced a lot of issues of people giving to each other and then having their
Eric Weiss (15:16.517)
Mmm.
Andy (15:31.566)
accounts frozen or just additional scrutiny on top of what they're doing there. And so that was hard for us. That was something that we needed to try to overcome with our partner there in PayPal. The other issue there was our user base, the audience we wanted to serve, they didn't have PayPal accounts. And so that was something that...
We had to coach them through. And for this specific use case, there was a lot of hurdles that they had to jump through. They had to go off of our site, start a PayPal account. One of the things at the time, because of the regulation of money and the internet at the time, was you had to give your social security number as part of that process, just upfront. And that is a very hard ask on the web as a product person. There's a lot of things we think about.
What are hard asks? What are things that you would ask a user to do? Social security number through a web page is one of the hardest. I'll tell you that. And so there's a lot of things there. But then just generally in the world, transacting on the internet was still really new for a lot of people. Putting your credit card in online was still in the common ethos of folks that
Eric Weiss (16:25.511)
Mm-hmm.
Eric Weiss (16:31.333)
Yeah, yeah, yeah.
Andy (16:49.1)
that thing that I shouldn't do. I should not give my credit card number out of, and online. And yeah. And so my, try to remember kind of where everyone was at the time because now like my browser just auto fills credit cards into everything and it's fine.
Eric Weiss (16:50.939)
Right. To some new startup. Yeah.
Eric Weiss (17:04.379)
Right. honestly, did you have, do you memorize your credit card numbers?
Andy (17:10.708)
back in the day, for sure. I knew all I knew my business credit card and my personal one by heart because I typed it in so many times into into things.
Eric Weiss (17:13.286)
Right?
Eric Weiss (17:20.089)
Exactly. So I used to have all my credit cards memorized and I don't think I've done that in over a decade. And just the other day, a friend of mine rattling off his credit card. like, you still remember your credit card? Everything is just trusted out there in the cloud now. Yeah, great idea. So, so, so, you know, so obviously, you know, we're both, we're both product people here too. So I understand the amount of friction that goes into getting people through that user experience. Was there any kind of, did you approach it? Like from a
Andy (17:26.497)
Yeah.
Eric Weiss (17:48.897)
from a user experience perspective, how did you help guide them through that friction?
Andy (17:53.388)
Yeah. So one of the, one of the things we knew is we had, we had some big asks and they came all upfront and we didn't really have a way to separate that from the upfront onboarding process at the time. And so in the absence of being able to, to separate those chunks or do them later, we just really walked you through it a very step-by-step way of here's what's going to happen next. Don't freak out, but we're going to ask for.
Eric Weiss (18:01.543)
Mm-hmm.
Eric Weiss (18:21.477)
Yeah, okay.
Andy (18:23.118)
for a payment account, and that's going to entail these sorts of things. So don't worry when you see these pieces of information being asked for. We were very clear on all of those steps, as well as provided you links to link you directly into, for example, the PayPal product where you got to tweak some settings and things. But even so, it was tough for a user to get through that, especially a tech timid user doing this, probably for the...
Eric Weiss (18:46.577)
pain from the air.
Eric Weiss (18:51.493)
Yeah. So these are like all the things that you're not supposed to do when you're designing a user experience, right? It's like forcing people to make a huge scary investment upfront before they get any kind of reward on their investment. So the only thing you could do is just kind of tease it, right? The word it's going to be so great once you get through all of this. Well, was there any kind of incremental reward or anything that you can kind of help them, you know, along the way?
Andy (18:52.632)
first time in this sort of setting.
Andy (18:58.696)
Right,
Yes. Totally. Yeah.
That's right. Yeah. Yeah. And to that point.
Yeah, is really nothing we could do. There's a couple of things we needed. We needed you to set up your e-wallet on PayPal. We needed you to connect that wallet to GoFundMe. And then we needed you to do the normal ask of you have to write about what you're raising money for. And telling a story is really hard for folks too. And so on that one, at least we could help them along a little bit better. That was all of our...
Eric Weiss (19:36.177)
Yeah.
Andy (19:44.054)
all on our platform, all in our control. But at the same time, is a hard thing for someone to articulate through writing online when you're not used to this use case. And so to your point, there's a lot of things we did up front. We showed lots of great examples of what people are doing on the platform, on the homepage, when we had them. In the very earliest days, it was more aspirational. And then as soon as we started getting some traction, even just
Eric Weiss (19:46.758)
Mm-hmm.
Eric Weiss (19:54.193)
Yeah.
Andy (20:13.634)
tiny, tiny amounts of traction, you at least can show social proof of here's other folks that have gone through the process and made it through the other side. And you're going to be okay. You just got to trust the process here a little bit. But we found that, you know, the fintech part was hard. The telling my story is very hard. And then sharing, we could have a lot of control over how to make that.
Eric Weiss (20:22.405)
Mm-hmm. Mm-hmm.
Andy (20:40.064)
easier and easier. And so we spent a lot of time iterating there on what's the simplest way that I've shared it out to the people that it will be impactful to share it to. But I almost didn't think about it. And so how can we make it really easy to send it out to all my email contacts? How do I make it really easy to post it onto my Facebook page? And so that's what we thought about. And so we were constantly iterating on that onboarding process.
Eric Weiss (20:44.102)
Hmm.
Eric Weiss (20:58.8)
Yeah.
Eric Weiss (21:09.383)
Mm-hmm.
Andy (21:09.452)
But it was still really hard. so, but we did have some folks that would make it through in the early days. And we, you know, we spent a lot of time with them after they were successful and just celebrating what they had done, because we knew if you could make it through all of those hoops up front, you would start raising money as early as like the next five minutes. It was so immediate, the response from folks when they would put their appeal, especially out on Facebook.
That sort of thing was eye-opening for a lot of reasons. One, it was amazing to see how generous people are to their friends and family and the people that they know. Even on that friends of friends layer, as your campaign page kind of made it rounds through your network and then the network of networks that you have with your friends of friends, people were amazingly generous. But it was also incredible to see
during the day, like the work day, how many people were just on Facebook constantly. And that ping of like, I post something and immediately, you know, somebody's going to see it and interact with it. And so that was, that feedback loop was so quick that you really needed to be ready for it. And so that's why the onboarding process was so difficult. And we really couldn't shave any more thing, any more friction out of that process. And so one of the things we knew is,
Eric Weiss (22:21.904)
Yeah.
Andy (22:33.44)
if there was gonna be that much friction in starting the product, there can't be any other friction anywhere else. And so we needed to, as product professionals, just continually shave off those sharp corners of all the rest of the product so that everything else just flows really smoothly.
Eric Weiss (22:51.569)
So, okay, so eventually the world got comfortable with, you know, sharing all of their information online and having no privacy and then GoFundMe grew to over a billion dollars in giving, right? Huge, huge accomplishment. So what were the biggest challenges that you dealt with on the scaling side?
Andy (22:56.942)
Yeah.
Andy (23:08.098)
Yeah. So we figured out the FinTech piece. We were able to collaborate on a platform with our early partners. WePay, they actually were a competitor to us. And then we ended up collaborating on this thing. They basically told us, hey, when we grow up, we want to be a payments company. They had an incredible relationship with the underlying FinTech Rail. So we collaborated on that. That was four years in. And so four years in of just being
extremely lean. We were a bootstrapped company. And so we had to figure out how to keep the lights on that entire time. We had raised maybe about $4 million for our users. And so not nothing, but certainly not a big business yet. the weeks after we finally finished the collaboration with WePay, we put it live to a small percentage of traffic.
And we saw that thing, we were going to scale it up over time. They were still competing with us at the time. We were a little wary of just going all in on this new platform that we just built. We saw in 24 hours, it was so overwhelmingly better for people that we have to put it live 100 % today. And so that happened in December of 2011. And then January, 2012, we had raised $4 million for our users up until that point.
Eric Weiss (24:25.308)
Yeah.
Andy (24:35.214)
And our users raised a million dollars in that January. And so then we knew we were off to the races. And so then we took a hard look at how lean we were. We knew we wanted to stay lean. We were bootstrapped. And so it was just Brad and myself at the time in the company. but we knew seeing the week over week growth in that January, it was incredible. And so we're like, OK, this is this is now to the point where we need to bring in some other.
great people onto this project. And firstly, I needed another engineer to help me out. As a sole engineer at a company that's going up into the right, that was scary. And so I wanted to bring somebody else I'd worked with before. And so we brought in Justin to come help us. We also need in
Eric Weiss (25:04.871)
Mm-hmm.
Eric Weiss (25:16.23)
Yeah.
Andy (25:26.638)
and customer support professionals. And so we needed somebody who was going to be extremely empathetic to where our users were at and could build an organization around themselves. Brad and I had been doing all the customer support in between everything else. And we knew this was a higher touch business. This is something new. People had a lot of anxiety about the internet and money at the same time. And we needed to talk to our customers directly. And so we...
foresaw that was going to get quickly out of hand for just the two of us. And so that was the next piece of the puzzle we needed to bring in. so scaling was all around, how do we bring in team members at the right time for the right roles and keep our bootstrap trajectory? And so we had certainly budget constraints of who we could bring in. We also had this constraint of, this was a new concept.
for folks and we were a very mission driven company. And so we had this constraint of on all hires, you had to be empathetic, you had to understand the mission. And because it was so new in those early days, we really had to spend a lot of time in the interview process explaining what the mission was and how this helps folks. Later, at least like maybe somebody had heard of us before, where they could go and play with the product and see the ways we're helping folks. But in those early days, we really had to do a lot of
selling as part of that to explain the mission. And customer support folks, I think that's normal course of business generally, as you want to find people who are good writers and empathetic and understand the business well. For us, layering on that mission aspect was pretty natural in that interview process. For technical hires, that was a little different. That's a little different off the script of the normal engineering interview process that one does.
But that was important to us as well. And all of our engineers helped out on customer support. At different times, we cycled them in so that they would understand talking to users on the customer support side. We ended up rebranding that team as Customer Happiness. But the Customer Happiness folks, they contributed back to the product roadmap every week. And so we really had one good small team that was lean, but
Andy (27:48.866)
was serving a lot of functions at the same time.
Eric Weiss (27:51.963)
How big did your team get? How many people did you end up having to be leading?
Andy (27:56.782)
Yeah, so after we hit that massive inflection point and started growing the team, that first year in 2012, we grew from two to 12 people and then 28. And then when I left the day to day at the very tail end of 2015, we were about 70, 75 folks at the time. And so that was a good size org. We were, I would say a tight 75.
We probably needed about double that for what we were doing, but everybody was just cranking. And so we had a great organization.
Eric Weiss (28:24.775)
Mm-hmm.
Eric Weiss (28:34.023)
So what were your biggest challenges becoming a leader or evolving as a leader from just being a very, very small team to a org of that size?
Andy (28:42.766)
Yeah, that's a great one. Both Brad and I had experience with leading teams around those sizes before. And so from a just managerial perspective, we did have some experience and tools in our toolbox from that. What no one really had an experience doing is leading such a visible, consumer-oriented
product that, you know, when we left in the day to day in 2015 was a household brand. And so, you know, that was a new experience for us. That was certainly a new experience for the team. And there was a lot of things that we had to learn along the way of like, how do we act as a platform that's so visible out there and that people rely on? And so there was tons of learnings around that. This was a new concept for folks. so,
Eric Weiss (29:16.519)
Yeah.
Andy (29:39.778)
the question of like, what should GoFundMe be used for? What should crowdfunding be used for? We have a continually answer at every stage of the company.
Eric Weiss (29:39.857)
Yeah.
Eric Weiss (29:48.135)
What about for you personally? What were the kind of the resources of the support system or the mentor or coaching? What kind of helped you evolve as a leader?
Andy (29:59.694)
We, um, you know, we were, we were pretty heads down focused on the product and the team. And so we didn't really have a lot of connection to the outside to, to mentors. Um, and because we were bootstrapped, we didn't have VCs to look to or, know, their EIRs or programs that they could, uh, link us up with. Um, we did have a couple of, um, fun sort of meetups that we would go to in San Diego and, and the.
biggest one we went to was tech founders in San Diego. And we got to meet a lot of other founders in our kind of stage of life of company. And there was good commiseration there across like, you know, how hard it is to be in a growth company. we didn't have a big network. that was actually, that was something that was tough on us. And so we looked to books and so we read a lot.
Eric Weiss (30:33.244)
Mm-hmm.
Eric Weiss (30:41.041)
Yeah.
Andy (30:56.654)
Both Brad and I read a lot of books, but we also had a book club inside of GoFundMe. so the entire team was reading business books at the time. And so we read things like 21 Laws of Leadership, The Lean Startup, those types of books that where people had been there, done that. Good to Great was a great one. Enchantment we loved. And it was just about how other leaders thought about
Eric Weiss (30:57.297)
Mm-hmm.
Eric Weiss (31:18.843)
Mm-hmm. Mm-hmm.
Andy (31:24.142)
product, how they thought about teams, how they thought about leadership and to have everyone in the company reading those books at the same time. It gave us a common lexicon and narrative to draw from when we saw these sorts of experiences happening within the company. You have somebody could raise their hand and like, I think this is like when they talk about the law of the lid and 21 laws of leadership, or I think this is like that time.
in Zappos that they talk about in delivering happiness. And it gave us a good shorthand to talk about these sorts of things. And we could debate back and forth how close they were to those specific examples. And we're like, OK, well, what did they do to solve that and those examples in that book? And let's look through those together.
Eric Weiss (31:55.046)
Mm-hmm.
Eric Weiss (32:10.364)
Yeah.
I've actually been, been trying to get a, founder's book club together, just kind cause I meet people out at the networking events. It's something that I used to do a long time ago when we were in like a coworking space and you know, the networks were a little bit tighter, but, yeah, I don't know. Maybe I'll ping you if we actually get it off the ground.
Andy (32:21.656)
Mm-hmm.
Andy (32:29.42)
Yeah. Yeah. And we find a lot of value just bringing our team together a lot. We had a weekly all hands and we were extraordinarily transparent about our metrics, where we were, who we're bringing on, how the team is growing. And so that we found a lot of wisdom within the team as well. When we had problems as leader, we would put it out to the team and get some guidance from the team as well, because they're out there talking to the customers, building the product.
Eric Weiss (32:47.461)
Mm-hmm.
Andy (32:58.466)
there was a lot of things that we were able to just collaborate on as a team.
Eric Weiss (33:03.397)
Yeah. Nice. Good. Well, at least you could circle the wagons. You know, I mean, it's something that, unfortunately I don't, it's not necessarily peculiar to San Diego, but it's something I know a lot of San Diego founders face and part of my work, you know, with startup San Diego and just as a general community builder, it's just easier when there's a tighter knit community. And so there's a lot of us that, you know, are trying really hard to bring founders together, but good for you for making it work, right. Bringing it, bringing it up internally.
Andy (33:07.32)
Yeah.
Andy (33:25.048)
Yeah.
Andy (33:30.798)
Yeah, absolutely. And that's something I believe in too. so after I left the day to day, since you're really 2016 onward, I've been giving a lot of time back into the San Diego ecosystem because I want to, I want to see that happen. And, and we are in a much different place now than we were, you know, when we started in 2008, certainly, but even from 2016, is night and day to where
Eric Weiss (33:45.703)
Yeah, I appreciate it.
Andy (33:57.154)
the support systems for founders are in the community and the way that you can connect folks. And so I really like that and I'm happy to give my time to that mission as well.
Eric Weiss (33:59.761)
Yeah.
I love it.
Eric Weiss (34:08.517)
I appreciate it. Yeah, definitely appreciate your time. so let's fast forward a little bit, right? So, you know, some years go by, you're, you're, you know, you're a serial entrepreneur and AI is blowing up the AI revolution is here. so what was the opportunity that you saw?
Andy (34:25.454)
So I started another company with Brad Damphouse. That didn't make it through the COVID lockdown. It was all around local services. But we really loved that product and what we were trying to build there. then interspaced with all of that, I've been mentoring at pretty much every incubator, accelerator.
Eric Weiss (34:34.875)
Mm.
Andy (34:50.188)
you know, that startup meetup in town. so Evo Nexus, Ad Astra, Connect, know, Techstars anywhere. And so one of the things that I fell into is an advisorship at Launch Factory. And so that's a venture studio here in town. And actually, we're physically located right now. And so we...
Eric Weiss (35:09.223)
.
Andy (35:16.64)
My role there was to be a product advisor for these new startups that the Venture Studio is creating. And one of the cohorts that they had was all around computer vision type solutions. And so I had a strong interest in AI. I built some early neural networks at the time. this was, let's call it, just after that COVID lockdown era, so in the 2020, 2021.
sort of timeframe. And so I was doing some more hobbyist learning at home, but I had the opportunity to interface with these companies that have been created.
Eric Weiss (35:45.254)
Mm-hmm.
Eric Weiss (35:56.264)
So some people, a lot of people were making sourdough bread. You were making robots see things. Got it.
Andy (36:00.077)
Yeah.
Yeah, and so it was really fun to see how they thought about these problems. And coming from a really traditional software development background, some of the things that they did when they were starting to build solutions to these problems, one, they were all different. And these are four companies that, yes, they have very different industries, but they were really solving the same fundamental problem. How do I find, measure, or count things in images or videos?
And to see them start so fundamentally different on those problems, that was a red flag to me to start with. And so I wanted to definitely investigate why that was happening there. The other thing that I saw there was that they were starting from really fundamental technologies. it was our, my co-founder for iPop.ai and CTO, Torsted, he likes to talk about, it's like,
As if you went to your investors and say, you know, I'm going to build like this e-commerce product and the first six months out of the, of time, we're just going to spend building the database layer of like, and we're not going to use, you know, MySQL. We're going to go from scratch. We're going to start writing like B-trees and figuring out indices and things like that. And so it was that level of granularity that people were starting with. And it was weird. It's like, why is this not a platform?
Eric Weiss (37:10.278)
Yeah.
Eric Weiss (37:18.495)
huh.
Andy (37:27.733)
And why is there not?
Eric Weiss (37:28.095)
I call this the apple pie problem. So there's a really great Carl Sagan quote. If you want to bake an apple pie from scratch, you first have to invent the universe. I use it all the time. What I see founders that do this. They're like, we have to build some incredible off, you know, not off the shelf, like from scratch technology. I'm like, dude, you're trying to bake an apple pie and you're inventing the universe first. Stop it.
Andy (37:33.431)
Yeah.
Andy (37:37.835)
There you go. Yeah, exactly. Yeah.
Don't know.
Andy (37:45.965)
Yes.
Andy (37:51.022)
Yeah. And there was no other way to do it at the time than to invent the universe. so that also raised a red flag. So there was just a lot of things in the fragmentation of that, of seeing those companies start that just really resonated with us as a problem. I met, so Brad Chisholm, my co-founder and now CEO of iPop, he was running Launch Factory at the time.
Eric Weiss (37:57.241)
Yeah.
Andy (38:19.63)
we brought in as a CTO advisor. And so we were all talking about this problem for months and looking at how these companies were solving these problems. And we're like, there's gotta be some fundamental technologies outside of the AI part. The AI part is hard and interesting and there's a lot of things that you can do with it. But there's an operational component to that as well that...
shouldn't be as hard, that should be more standard. And there's a lot of things we could do to bring in this flexible operational component to place these computer vision models in and train them and operate them in the wild in the ways that need to scale and need to be delivered in different modalities. And so we started thinking about what would that company be, what would it look like? And we thought it would look like a platform where you could come and use
pretty much web-style SDKs to access this sort of AI. And this is before the common, or before chat GBT was very common for folks. And so we were thinking about this in an interesting vacuum. And then right after that, as we saw the proliferation of LLMs, it gave us some more to think about of how people want to interact with these technologies.
Eric Weiss (39:44.997)
Yeah. So is this, there, are there particular sort of industries or use cases where this works best? this like a hot dog, not hot dog kind of situation?
Andy (39:54.304)
Yeah. So, so I think one of the fun parts about computer vision is, helping folks get over the hurdles. And I know I'm there too, right? Of like, think about your product or technology. There are either users giving you images or videos. There's live stream being created by your cameras. We're all constantly on zoom. There's so much visual data all around us in our business that's unstructured.
that we're not pulling insights out of, that we're not using for interesting product experiences, that we're not using as guardrails around our product experiences. And we treat those things as black boxes right now. When I take a profile photo from my users, I may have something that helps me check that it's safe for work kind of thing. But really, I throw it in a bucket. And if it was all black pixels, I wouldn't know the difference.
And there's a wealth of data our users are trying to tell us in that. There's a wealth of data your customers are telling you just by showing up in your retail store or on the job at an industrial setting that you can tell from cameras that there's so much data there. And so we're not using that data because it's really hard to structure. And so
Eric Weiss (41:15.879)
Mm-hmm.
Andy (41:16.928)
Every business has this unstructured visual data that they should be using. And so I think it's our job to be really horizontal as a platform to help folks understand for every industry, how do you structure that data and how to use it for driving business goals. And so we want to keep the platform extremely horizontal. We have customers in lots of different industries and they're very varied. And there are some industries that, yeah.
Eric Weiss (41:42.715)
Yeah, what are some cool examples right now?
Andy (41:46.156)
Yeah, so we have a customer that is taking video that users are taking of themselves after watching some coaching around pickleball. And it's helping them analyze their pickleball swing and how their stance is and how they're holding the paddle. So you have that kind of experience on the consumer side of things. Then you have one of our customers sends couriers out to
medical labs to pick up medical lab samples. They are using us as a QA in their workflow process. Those couriers can take a picture of the samples that they are picking up. It can take the metadata that we're structuring, how many samples, what typology of those samples, what does it say on the sample bag. That can be used to ping their database and make sure these are the right samples. Or, hey, you have two of the right samples, but you're supposed to pick up a third. And just from the photo,
knowing that that QA process is happening is adding an extra component to their workflow and accuracy of how they do business. We have customers that are using us in finding the types of material that are coming through in their factories. so there's more industrial settings, like lot of industrial warehouse uses of looking at cameras.
Eric Weiss (43:12.241)
Yeah.
Andy (43:14.03)
seeing how cargo passes through the warehouse, seeing how different materials pass through an assembly line. Those are the types of problems that we're solving right now. And aside from the industrial side, I think they're actually outside of the normal spectrum of what people think about for computer vision traditionally. And so I think there's a new wave here that we're
Eric Weiss (43:35.345)
Yeah.
Andy (43:40.46)
that we're writing where people are starting to think about how do I use cameras for data.
Eric Weiss (43:44.487)
So now are your customers, I imagine are developers, right? They're other founders or app developers that are building. So they are technically savvy, but what are some of the product and UX considerations that you have here?
Andy (44:00.098)
Yeah, so we have two customers. Primarily our customers are developers, although we have a lot of business folks that see the capability of what we can do, and they're investigating using iPop as well. And so if there are ways that we can deliver our product to non-technical customers as well, we're exploring those. have, so I guess on the developer side,
One of the things that we have to do is we still have to make it accessible. You have to be able to sign up and get to value very quickly. And so we try to make that value right at the forefront. As soon as you sign in, you have a default endpoint that you can start using right away. Everything has to be SDK first. And so everything you can do through our UI, you have to be able to do through the SDK as well. And so we've really co-developed the product and the SDK product at the same time.
Eric Weiss (44:51.717)
Mm.
Andy (44:57.422)
to allow for that. And then, like I mentioned, we want to be able to serve customers who are not developers. And so one of our newest products, Reports, has a workflow where to create a visual intelligence report, you're giving us a batch of images that are related in some way. You can ask a series of questions about that, and we deliver to you a CSV file and a PDF that represents that data. And so an example there is
One of the workshops we just did recently with an industry is around the aerial inspection space, looking at people's roofs and other things with drone footage. And there's a workflow right now where you go out and take bunch of drone footage, but then an expert has to sit there and look at the footage or look at all those photos. And it takes a lot of expert time to discover things.
This use case that we talked through on our workshop is how some of our users are using our visual intelligence reports right now by taking those drone photos for a particular house, throwing them all into the same report, and asking questions like, did you find missing shingles? Did you find damage like hail or other sorts of storm damage to the roof material? What material is it?
Is there HVAC up there? Are there solar panels up there? What's the condition of those types of things? And so just by asking prompt-based questions, you can come up with this big corpus of across a lot of images very quickly. And it saves a lot of time on the expert side to roll up that data just from a batch of visual data, images and video.
Eric Weiss (46:28.667)
Hmm.
Eric Weiss (46:45.083)
But I didn't know. like I can just upload a hundred pictures of roofs and it knows what HVAC is and it knows what a crack tile is without me having to tell it. Interesting.
Andy (46:53.452)
Yes. Yeah, absolutely. And so that's the strength of these modern DLMs, the ability to take these common questions, anything you would naturally find out there on the internet as visual sorts of problems that people are thinking about is trained on. Yeah. Yeah.
Eric Weiss (47:07.599)
Okay, okay. So it's trained on the internet, public data. Okay, so I don't have to train the model. I just feed it and say, tell me what you can see with the stuff I'm giving you, but it already knows what everything is based on.
Andy (47:21.442)
Correct. And there are definitely specialized use cases where it's not great at finding things yet. And those are types of, usually types of imagery you don't normally see out there in the world, like medical imagery and more specialized diagrams and things like that. But the great part is these models are fine-tunable. And so we can give them more data around a specific industry and have them learn from experts. And so one of the things that we've created
in iPop is a self-service training platform that allows people to bring in images and interact with foundational models to auto-label those images and give their AI more data around their specific use case and to fine tune to their customer.
Eric Weiss (48:11.921)
wow. So cool. Okay. So, so, you know, you're, you're several startups in here. what are some things that you're thinking of that you're like, I'm not going to make that mistake again. Like what are some new approaches that you're taking this time around?
Andy (48:16.92)
Yeah.
Andy (48:24.206)
For this one, we bootstrapped GoFundMe, and I would do that again. That was the right decision for that particular product. For this one, AI scales a little bit different than web technologies, and also the team scales differently. We have some really high-paid experts in our team, and that's different than how we scale out a web product. And so this time around, it made sense to do
more of a traditional VC path where we had a friends and family round, we had a seed round that we closed earlier this year, and right now we're thinking about how do we hit the milestones we need to start raising a Series A next year. And so that's a, I don't know if that's a learning, but it's definitely a difference of how we're starting this one, and I think that's the right path for this particular company. I think one of the learnings that I'm keeping is,
Keep it as lean as possible. Smaller teams, if we can do the work, it's better for communication, it's better for speed. Speed is everything in the AI business, especially. There's competitors popping up all the time for lots of different things, but there's also just so much noise within the AI space and new models coming out all the time. And so that speed is so critical to stay up.
Eric Weiss (49:24.593)
Mm-hmm.
Eric Weiss (49:33.275)
Mm-hmm. Mm-hmm.
Andy (49:52.078)
ahead of where you need to be and also stay aware of what's happening out there. And when there are new technologies, you really have to figure out a way to very quickly bring them into your platform.
Eric Weiss (50:03.015)
Wow. Any advice that you would give to other startup founders in our audience who are, maybe this is their first startup.
Andy (50:13.006)
Yeah. Yeah. You know, I think that the advice that I generally have for startup founders is it is going to be hard. It's always hard. You know, this is I believe, you know, my eighth or ninth company that I've been a part of in the startup community. And it's never easy. We I think we hear these kind of whitewash tales of like Instagram, you know, 17 months later, their billion dollar company and kind of expect it to go that way.
Eric Weiss (50:32.891)
Mm-hmm.
Andy (50:43.238)
when you haven't done it before and it's, I've never seen that in firsthand, it's, it's always a struggle and you're going to be tested. and so one of the things that I always coach founders that I work with on is if you're trying to go it, alone, see if you can find somebody with, dissimilar skill sets that also believes in this mission and have that synergy of, of, know, different expert areas that.
Eric Weiss (50:46.631)
Yeah.
Andy (51:10.818)
the founding team can tackle, not just for the additional skills in the team and additional brain and focus thinking about those things, but also the emotional support of having a co-founder because of how hard it is. It's going to test you and having somebody that you're really going at this with together, then it just allows for you to have that emotional support system, which I think is really critical.
Eric Weiss (51:22.257)
Yeah.
Andy (51:40.206)
Not being surprised when it does go well, it actually gets harder, right? And so growing is even harder than not growing and being in that like nebulous space of is this a real company or not? When you're growing, you know you're a real company, but there's so much pressure on the leaders to continue the growth, to continue that trajectory. You're going to be meeting new obstacles you've never met before.
Eric Weiss (51:46.661)
You're right.
Eric Weiss (51:50.844)
Yeah.
Andy (52:09.12)
all the time that happens constantly. And so every plateau you break through is just the next problem and it's going to be more complex and you're going to need different skill sets to solve it. so bringing the team around those, those growth problems and it's going to stress out your team too. And so you're going to have to be there with a cool head to also keep your team calm, focused and productive during that, that stress or growth period. so.
Eric Weiss (52:10.523)
Yeah.
Eric Weiss (52:26.565)
Yeah.
Andy (52:37.558)
I think those are two of the things, like general advice I usually talk to teens about, but yeah.
Eric Weiss (52:41.167)
Yeah. Yeah. It's so fun. I I couldn't, I couldn't agree more. So I I've been an executive coach now full time for almost a decade, but you know, for probably about 15 years as a mentor advisor and so on. And I find the same exact thing that especially with younger or first time founders, they always come to me looking for an easy button. They think that they're going to hire me. I've got all the answers and I'm going to make this easy for them. And what I have to
Andy (53:01.964)
Mm-hmm.
Eric Weiss (53:08.123)
teach them is that I'm not here to make it easy for you. I'm here to make you better at dealing with hard things. And so actually my, my sort of core leadership framework is actually based on stoicism. And so I actually built this whole framework, call it the stoic CEO, where I go through all of the core stoic principles and apply it to leadership in a modern context. And it really is just how do we, it's emotional regulation and how do we
Andy (53:15.63)
Yeah.
Andy (53:24.696)
Nice
Eric Weiss (53:34.395)
you know, focus on what we can control and not let ourselves worry about the future and deal, you know, worry about other people's opinions and, just kind of follow the thing that, that, drives us internally, intrinsically every day. And, and, and that just sort of gives us a, an even keel so that we can just continue to, deal with the hard things as they come and just realize that it's never going to be easy. If you want it to be easy, this is not the place for you.
Andy (54:00.366)
That's awesome. Yeah, I love stoic philosophy. I love the daily stoic and Ryan Holiday. I get a lot of solace out of reading through those problems across the ages. But I like how you wordsmith that of helping founders deal with hard things. You're not going to be able to make it necessarily easier, but you can make it more understandable about dealing with hard problems in the future. so, yeah, I like that a lot. think that the...
Eric Weiss (54:05.691)
Yeah, exactly.
Andy (54:27.244)
very well aligns with the folks at HectoCure. Love it. That's awesome.
Eric Weiss (54:29.319)
Well, I'll send you, I'll send you the link. actually have a manuscript for a book that maybe one day I'll finish, but I'll send it to you. all the exercises and everything. But, uh, well, Andy, know we're, we've been, uh, we're way over what I expected, but, it's always, you know, so great to talk to somebody like you. And it was so great to have you on the show today. Um, uh, yeah, I'm not, if I know if I ask you another question, we're just going to go on for another 15 minutes. It was so great to have you on the show. Andy, I wish you the most success. I thank you so much for your support for, know, for the startup community here in San Diego.
Andy (54:50.152)
Yeah.
Eric Weiss (54:58.073)
and I look forward to seeing a lot more of you.
Andy (55:00.952)
Fantastic. Well, thanks for having me. It's good talking with you.
Eric Weiss (55:03.602)
All right, thanks, Andy.