1 Dec 2020 · 55 min
In this hour‑long podcast, Jude Gomila recounts his entrepreneurial journey, starting with an engineering degree at Cambridge and early internships on a solar‑inverter startup and at Cambridge Sensors. He describes a university‑driven medical device project (ConnectEase) that failed to secure funding, followed by a series of experimental ventures – egg‑packaging imports, a digital photo‑frame brand sold to Harrods and Selfridges, and a short‑term Nintendo Wii arbitrage business. A summer stint in Silicon Valley led him to co‑found Heyzap, get into Y Combinator, and raise a seed round led by Union Square Ventures with board members Naval Ravikant, Albert Wenger and Chris Dixon. He reflects on the hardships of fundraising during the 2009 crash, the lessons about not falling in love with a single deal, and the importance of hustle and frugality. Gomila then shifts to his role as an angel investor, describing his rolling fund, the evolution of his note‑taking and scoring system, the importance of founder chemistry versus market metrics, and how he has backed over 200 companies, including ten unicorns. He shares his perspective on crypto as a new landscape for decentralized business models, and his ongoing curiosity about biology and space. The conversation culminates with an overview of his current venture, Golden (formerly Golgon), a large‑scale knowledge‑graph platform that aggregates entities (companies, products, technologies, people) from disparate public sources using modern NLP (entity extraction, linking, event detection, NLG). Backed by Andreessen Horowitz, Founders Fund and others, Golden aims to provide a queryable, open‑licensed database that surpasses Wikipedia and search engines in precision and coverage, with human‑in‑the‑loop AI suggestions to ensure data quality. Gomila explains why advances in LLMs, cheaper compute, and accumulated domain expertise make now the right time to attempt this ambitious project.
Chapters
Thank you so much, Jude, for coming on the show today. I really appreciate you taking the time. I've been following your angel investing for some time, and I'm excited to talk about your company, Golden, today. For those people who aren't so familiar, would you mind just starting by talking about Your story and how you got to where you are today?
Sure. Yeah. So there's a bit of a long story and there's a short story.
We always like the long story on this podcast.
Okay. I'll try and get something in between because I think it's actually quite— entrepreneurial stories are always really interesting. Like there's so much detail that people can go into about each step, especially things that they did in the past that didn't actually work or scale or failed and like how they got there. So I'll give a somewhat brief story. So I'm originally from the UK and I went and did engineering at Cambridge University. And I had this inkling that I didn't want to really work for companies straight off the bat, that there were things that just felt like stale and old and wrong and slow about these companies, especially the larger companies. So I was highly interested in innovation and new things happening and things being done in engineering that were like, wow, that's pretty cool. You know, nanotech and biology and all sorts of different kinds of fields. So I started— I actually did some internships in my second year And the first year was actually at a solar power company. And that was a startup in the basement of Cambridge University Engineering Department. And we were working on inverters. Sorry, inverters. Inverter story is a different story. And they effectively, you know, there's 3 people, PhDs in Cambridge Engineering Department, started a company up to build scalable solar panels. that could all connect together and work as a network and not have bad effects. So I worked with them. I had no idea how to even work in a workplace. They put me on an impossible problem, almost zero management. They would check in on me on Monday morning and then Friday, and there was no team interaction. So it was just me and the screen.
And this is around what time you're working on Solar?
Yeah, this is 2012. Yeah, 2002. So 18 years ago.
So you're tasked with figuring out solar by yourself in 2002, basically.
Figuring out a problem in the boxes that would invert the DC to the AC. And we had this bottleneck in the industry. They didn't tell me it was an industry problem. And they just said like, hey, you've got to solve this problem. We've got this problem. The capacitor keeps failing. Can you look into it? So I spent a lot of time looking into that capacitor issue. And finally realizing that everybody had the problem and it kept failing. And it was one of the reasons that this inverter kept failing across the industry. So that was an interesting awakening for me because I thought I was going to go in there and be helping build the company. And I had no clue about any expectation settings or how to work with people. And so I just delved into this problem for ages in this freezing cold basement in the engineering department. Hopefully it was of some use to them. And then I looked into other kinds of problems as well. I don't think I solved it, but maybe compiled a bunch of information to help them solve the problem. Then I went and worked for another entrepreneur. And this person had built Cambridge Sensors. So he was a bit of a technical genius. Synaptics, who built the iPod touch dial technology, if you remember that. They were this electronics company in the Cambridge Research Park and I went and worked there. And I worked underneath the entrepreneur that built Cambridge Sensors and Synaptics had bought him out. And that was really interesting actually. That was like very high-grade people. They weren't quite a startup. They were still like kind of a reasonably established technology company. And that was really cool. We built like some really cool stuff. Like we tried to get the screen so you could use capacitive and inductive sensing at the same time, you know, for kind of touchscreens. So we got to build a touchscreen prototype, which is really fun. And I learned a lot about what, what, you know, more high-tech— the word was high-tech back then, you know, like tech startup. It wasn't really a startup, but, you know, company was doing and how, you know, it was working a bit with Apple. So then I got out of that and that was over the summer. And I started applying for competitions. And in Cambridge University Entrepreneurs, they had a competition for business ideas. So I applied for that. And the idea was to try and take an MP3 player, miniaturize it, and put it in a headphone. So you could play 100 tracks, have them on autopilot, put it in earphones, you don't have to connect it to anything. So kind of like AirPods, but it's got the tracks built into it. And this was cool. I think this is a cool idea. I didn't get anyone. I don't think I had anyone contact me back, which is ridiculous.
Do you remember how you came up with that idea?
I just thought that— well, actually, no, there is some backstory to that.
During—
pre-university, I was messing around with electronics. There was another guy doing the electronics course with me who was a bit of a genius and he had built an MP3 player. This was in 2000. Was it 2000? No, sorry, 2001. He built an MP3 player on a breadboard from scratch and it played about 4 seconds of audio and I was just so impressed with it. It was like the most complicated multi-breadboard, wires everywhere, like 5 layers thick. And I was just like, wow, that's really cool. So I asked him lots of questions about the MP3 player and how he made it. And I just thought about the user story of the user problem of wanting to play music while in the gym or running around. And there wasn't like— we weren't really dealing with smartphones here. So we had these bulky MP3 players. And I just thought, well, maybe I just want a couple of tracks and we can probably get that on a small amount of memory. and get it into the ear, into the kind of like thing that would be inside your ear and a little bit of battery for like a run or something. So anyway, it went nowhere. It was just like a kind of write-up of an idea and no one got back to me. That's fine. And then it comes, then I went to third year and we, in our course, we were actually allowed to go and try and make a business, like a virtual company where it's like, you're not making it for real, but you're treating it as if you're making it for real, which is very strange. It's like, why not do the real thing? So we went and interviewed— we actually chose a problem. One of our team members' dad had proteinitis— sorry, kidney issues, and he was doing peritoneal dialysis to fix it, but they had all these problems. So we went and interviewed all these different patients from Addenbrooke's, which was a local hospital. And that was kind of really interesting because we were getting a bit closer to building a real company here. It was still a university exercise for our coursework. And I took it really seriously and I wanted to turn it into a real company. So I got— you're getting actually an interesting story here because I don't think I've gone into this detail on other podcasts. So I got the CEO of Baxter Healthcare UK to be my mentor. We made the prototypes in Japan because one of the team members, her dad's family owned the largest, like one of the most precise machine tool companies in Japan. So we actually started making this thing for real. We had 2 really talented people on the team in terms of design. So actually, we built like a pretty decent device for improving peritoneal dialysis. The company was ConnectEase. The brand was pretty cool. And I went for this competition. Cambridge University had a £50K competition. You would win £50,000, actually, if you won the competition. So I spent all year writing the business plan and really trying to push it forward and way beyond the coursework and really taking it kind of seriously, wanting to actually do it as a spin-out. The other team just wanted— the rest of the team just wanted to do it as coursework. And so we straight off the bat had problems possibly with the IP because it was split between 4 people and maybe the university. Really, it wasn't like their cup of tea. They wanted to go into companies and that's fine.
So it wasn't even supposed to be a real company, right? It wasn't supposed to be a real company.
It was just a project that we had addressed real user problems. It was a giant market with consumable model and people were interested in actually taking it forward when we pitched them and showed them it. And the prototypes actually were pretty good. They were much better than the competition and the Baxter Healthcare CEO was like, this is interesting. So I was trying to push a company forward here without any kind of— it wasn't the right situation where you've got co-founders together that want to do something. It's really a university project that I just wanted to take it to market for real. And I applied to this competition and the other people helped. We're moving ahead a little bit because they were interested in the competition and that having implications for resumes. If you win a competition, it's good for your resume, not necessarily doing the actual company. And so we went forward a little bit with this and then we went all the way to the final and we were the only non-PhDs to get through to the final. Because this was normally for serious companies that were going to spin out like Alstom, the chemical smell sensor company from Cambridge, and other ones that were real companies trying to spin out. So we were the only non-PhDs. So we were the youngest group, the undergraduates or whatever. And up against all these PhDs that were doing very technical companies. And then they told us like, hey, there's no money this year. Like the sponsors pulled out the night before all these presentations were supposed to happen on stage in front of people in the university. And that was really disappointing. And there was no, absolutely no funding. And it was like, well, can we have done all this work for no funding? And we have to also lay out that the angel investment community was not really there at all in the UK at this time. So there wasn't any other ways to get money and get this off the ground. So this project was kind of dead in the water at that point. It's funny, people even did the presentations on stage pitching for the $50,000, everybody knowing that there's no actual money, just for the audience. It was kind of a farce. I was super pissed off. I really wanted to make this happen. It was kind of disappointing. It could still happen, actually. Someone could build this still. So, you know, move beyond that. And then in the 4th year, got together with 2 other people that wanted to do a company and we started trying different things out and that was cool. So these were my first formal co-founders. And we started just making a holding company to try out different things. And we actually experimented with lots of fun things. So we ended up having something like 4% of the UK market share for egg packaging. So we started importing egg packaging from China. We noticed an opportunity for cost reduction and got it into the supermarkets and expanded an Asian manufacturer into Europe. And we made the company for them and all this kind of stuff. We didn't know what we were doing. It was a really terrible business. Really hard to get the sales done with hardly any competitive advantage on the product side and no margin for us. So we learned pretty quickly, wrapped that up, got out of that. And then we started thinking, hey, what would be a good thing to do for Christmas? what would be a good thing to sell? And we came across— we looked through lots of different electronics products and we came across digital photo frames. So we got a feeling that this was going to be a hit for Christmas and it actually turned out to be true. So we started OEMing, building this digital photo frame company, and we went to China and we flew into China quite a few times to investigate the manufacturing plants and find the right SKUs. And we found this good one that was actually a pretty pretty market-leading in the UK. So we formed a brand around it. And then we had to get sales. So we managed to break in and sell this product into Harrods and Selfridges, which are very prestigious stores in the UK. And we actually had— we were very lucky to be able to position this as like the best frame in the UK, this digital photo frame in the UK. And then the prices, then natural electronics and monetization started happening and other people started getting in. And the prices started collapsing on us. And actually, we had to get out of that. And at that point, one of the team members had a visa issue and he needed to focus on getting his visa sorted by joining a company. And the other one wanted to go into software because web companies were getting super hot and interesting. So we decided to disband. And I took a I had not been paid for a long time. So I'd done 18 months with no cash. And we're putting everything back in the business. So I spent 76 hours trying to come up with a way to make money. And it kind of worked, actually. So I found an opportunity to buy Nintendo Wiis in the UK. They were running out of stock. This is actually happening with PlayStations right now. I think PlayStations are out of stock. So I saw this coming and I bought a bunch of them. And I didn't have any money, so I had to be really creative in terms of getting capital. And then I ended up selling them, setting up a website for it, and just being really nifty on Google AdWords, winning the ecommerce battle, and supplying these Nintendo Wiis. And this is a very short-term opportunity that worked for me, and it made some money. So I managed to So that, you know, one of my friends had actually flown, moved out to SF to join a YC company. And, you know, he said, hey, why don't you come over to San Francisco and do like a summer work with us over the summer? So I went and did that. And that was really interesting to be immersed into an early YC community of say 100 YC people all doing really, really quite wild companies. Like real interesting companies like Dropbox and Discuss and people were building stuff. So I went and dropped into that company and realized that that company was not doing anything interesting. It wasn't working. So I went solo and started looking around for co-founders, interviewing different people, hanging out in the Valley and looking at different companies and learning about things. And then decided to actually co-found with a person that had come out already. And we went and formed Heyzap. And that, you know, we wanted to get into Y Combinator. We needed something as well to ground our visas. So we started working furiously on getting distribution traction. And it kind of worked. So we got into YC. And then we struggled with cash flow because we actually YC at the time didn't give that much money out, $30K or something. So after we paid for our visas, $20K into the visas, we had $10K left, and we hadn't even started YC yet. So we went negative into our bank balances and into our credit card debt. And we had some form of traction. We had people installing this widget. So we compiled all the Flash games out there, and people were installing this on the websites, and we had 300,000 websites that installed it. Off the back of that, we actually managed to land Union Square Ventures as a lead investor. We got backed by really tier A VC. From that point, we started to learn the finer points of scalable companies. We got Naval and Albert Wenger to join our board, and we learned from them for 4 or 5 years. Then we were getting really into the heart of fundraising in April 2019. '09 was really hard because the market crash had happened. And we had to learn the hard way of how to raise money, what is traction, what is a good business. And we were watching some companies in our batch do phenomenally well like Airbnb. They were in the same batch as us in 2009. So yeah, then we built our pays app, we got profitable, and we ended up selling at $45 million in 2016. So that was the journey. And then I went straight into my next company, Golden, having learned the hard way of failing on different things and some things having little bits of sparks of success and seeing different permutations of models, some being better than others and seeing other YC companies doing interesting things. That's the kind of medium-scale version of the answer.
No, it's an awesome story. And I definitely want to talk a lot about Golden hopefully today, but I want to drill in on a couple aspects of the earlier story because I think it's super fascinating. And obviously there's been this thread of you wanted to work for yourself and do your own thing and build something that was super interesting from the very early days. And there might have been a couple of schemes that were more so focused on just Making a quick buck so that you could enable yourself to pursue these greater goals like the egg business, the egg packaging business.
And yeah, that was a quick one compared to say Connectese, like fixing peritoneal dialysis as a huge medical problem. And like I kept dancing between like this not having any cash and there not being any ecosystem to really raise money for people that didn't know what they were doing and didn't have PhDs.
Yeah. And I can sympathize. I mean, not from my own experience, but it sounds awful that they would put you on for a year that they had this $50,000 reward and then take it off the table and you still have to present like you're going for it.
I learned a big lesson as well. It was just like you absolutely— and Naval like said this again many years later, you cannot fall in— actually, maybe Paul Graham said this, you cannot fall in love with a deal. So you cannot just have one deal that you're relying on. It's absolutely crazy to do that. And you can't get obsessed with anything really, because it could fall through. So you need a pipeline and you need something else. And also, the other meta lesson there is that what you're building needs to be the magnet, because if you really did have something super, super good, you probably will have money coming at you. Competitions are a really bad one to go after. There's only one slot to win. Whereas raising a normal— raising much like what we're raising today is very trivial compared to before. It really was difficult before. And now it's extremely easy to raise money, which is a great thing in ways, and it means that people can try more experiments out with companies as well. Yeah.
Yeah, so speaking of competition, you, in the very early days of YC, identified it as a really attractive opportunity, saw things like Dropbox being built, ended up joining in the same batch as Airbnb. How do you think you were able to make that? That wasn't obvious back then that YC would be what it's become today. How did you see that and get excited then, and then ultimately kind of create your own fate in terms of being able to get in with Heyzap?
It seemed like the nexus of thinking in terms of these people are extremely smart. They are smart, they're serious, and they're some of the most interesting characters I bumped into. You have the James Lindenbaums of Heroku, who was schooling Imad and I about how to raise and how to raise VC and do it correctly. And just all these different characters. Dropbox gave us a desk for, I don't know how long it was. It felt like months. It wasn't necessarily that. It was maybe probably 2 months or something. So we were hanging out, working on the back of our startup when they were 6 people, and they're like, they're just in there all the time, like working. They started late, but they worked all the time. And then they go back to Guitar Hero or whatever, and then back into work. And they did that until the very late night. And it just felt that was a real startup. And seeing that was kind of cool. And seeing other ones, you know, we kind of like, we didn't even have an office for the early Heyzap days. We just jumped into other people's offices, and they kind of let us in, which is cool. So yeah, there was obviously like Paul Graham being a central figure. And this was the place to go to, to really learn about how to build companies. And there was the money around it and the Demo Days, the learnings. And it was quite cool, right? It wasn't everywhere. It was in one particular place and the connection to Hacker News. So it really was pinpointed in one place, and we just knew this was the place to get in. And I'd set my sights, I want to get into YC, I want to learn how to build startups and learn about the ecosystem.
Yeah, I want to zoom in on something you said earlier as well. You talked about the digital frames business that you'd started prior, and that was reasonably successful, bringing the Chinese product to Europe. And you mentioned how you had a reasonably successful time selling it At the higher end of the market. And then I want to connect that. I don't know, you'll tell me whether this is fair or not, but later on as you're fundraising, you go ahead and you're able to get USV to lead the round at a time that's very difficult to raise money in.
It was so hard. We couldn't get— we toured the Valley to try and raise money and we went to every single angel. And we just couldn't— no one had any money. Nobody wanted to do anything because of the crash. And it was just like the worst time to— this was the time where Sequoia said the good times are over. The email went out. It was just really miserable. And half the batch died just not being able to raise money, which we thought was standard back then, by the way. And now it doesn't have to be like that. So it was extreme power law on raising on Demo Day. And so we went and pitched. We went and set up those kind of 6, 7 meetings per day and did the hard work and met with maybe 170 investing angels or groups. And we had to learn the dynamics. We learned the dynamics on the job, right? And what was interesting is that we kept getting close. We kept resonating with the smarter investors. And that was a good thing. So we ended up closing the best ones we were pitching. And that was kind of counterintuitive. And for USV, I remember, like, we were so strict on not— and by the way, people weren't building, people weren't doing high-resolution fundraising really that much. So we had to get like a round together. There had to be a lead. So it was even harder before that you had to— all the angels expected a lead. So you had to have a VC lead it really to get it together. And the syndicated rounds were just starting to break through. I'm not talking about angelless syndicates, just syndicated rounds. So that was really hard to break the chicken and egg problem of the raise, because you couldn't just take a $50K here, $100K here. That was happening for some of the best startups. It may have just been like where we were in the middle ground or upper middle ground. So yeah, we did all these crazy meetings all the time, and we really had some horror story pitches, not necessarily from our side, but what we saw on the other side. We saw a lot of people that were asking for ridiculous things like board seats for $50K and crazy provisions, all the kind of nasty stuff that VCs used to ask for in the old days before things went vanilla and safe notes and all this stuff popped up. But we really liked the Union Square Ventures people and we didn't have— we were really running low. It was like our last $600 to get to New York and we just thought, okay, let's just do it. We had so many nos, we're like, well, let's just roll the dice and go to New York and do it. It was pretty tough doing the pitch and having no sleep. And no sleep the day before the pitch, really, because of the budgets we were on. And I remember Iman and I were really trying to save cash and being super scrappy. And we were working really hard as well. So that was— but yeah, it worked. And we got a deal. term sheet from them. And it was great that they took a bet on us. And that was the kind of big break, I think. Getting to YC and getting Union Square Ventures was a big break.
What was it like working with Albert and Naval? I think Chris Dixon also invested back then. These are people who obviously are very well regarded as investors today. I'm not sure that it was such the case back then, not that they were I'm sure well regarded, but all 3 of them, I'd say it's safe to say they've really built their resume over the last 10 years. What was it like, those early conversations?
Really good. Yeah.
Board meeting or otherwise?
I think people were figuring out a lot of things. The playbook was not— I mean, I'm not saying the playbook's understood for startups. I think it is a lot more now and a lot more down to execution. The playbooks weren't known so well. So people were figuring things out. And those conversations are really, really fruitful in terms of figuring that out. Crypto was kind of kicking off as well. So we were having conversations about crypto after the board meetings were finished between the 4 of us. And Union Square Ventures and obviously Naval played a big part in crypto. And so the 4 of us were having all these intellectual conversations about network effects. crypto incentives beyond the company just being built. And that was really cool to have those 4 people in the room on a regular basis and just discussing things. And I remember with Chris Dixon, we met him in New York on one of our side meetings to see if we could get him to invest in us. And I think he was working— yeah, he was working on his startup that got bought by eBay. So even in the roster, we had pitched Shervin. Shervin was really useful. He was at, I think he was at Menlo at the time. So there were lots of interesting people that are now, have grown into bigger roles and stuff like that in the ecosystem that were along the route that we got help from. And the Valley was smaller back then, for sure. People like Loïc, them, like signed our O-1 visa letters. We had a visa problem. We had to get a visa sorted out as well. So that was a kind of special mission in itself. So I would say it was like super hustling with no leverage. And we managed to kind of break through that. And I think it's— and then getting into more refined execution, understanding things and getting better and better. So it's still, it's a journey, right? Like it takes, you know, 10 years. Some people can figure it out on day one, or they're either just naturally brilliant or they run into something really good. I think everybody else has to go through a kind of journey to get there as well.
You've talked about running on zero a lot and running your bank accounts dry and really hustling. I understand even when you did start to make some money, you kept allocating and allocating it towards angel investments, and now you've gotten a couple hundred over the last decade or so. What keeps you draining the bank account to get more angel investments out there? Is it just an obsession with supporting these early-stage ventures, or you just don't really care about the money, or what's your thinking behind that?
I mean, certainly an obsession with the startups and the problems that are being solved and the kind of engine of creativity that happens at the start of company formation. I am definitely also, because it means so for such a long time being comfortable with operating on zero or negative that it doesn't faze me. It's like, it doesn't, I don't care. And so that means that I see all that extra stuff coming in as, and once you're frugal for a while, you're frugal anyway. So you get some cash coming in and you're like, okay, well, you're not gonna suddenly, some people do change it. They're like, okay, I'm gonna buy an expensive car and I'm gonna buy an expensive house and I'm going to change the way I live and upgrade my clothes. And I didn't take that path. So I was just like, okay, it's going back into startups. And this is what I want to get engaged with and this is what I enjoy and this is the engine for change for the future. And this is the engine for learning as well. Learning how to be a better CEO, learning about technology, best practices, and just go creating stuff. And I remember Eddie Kim, the CTO of Gusto, He was— correct me if I'm wrong, but I think he had done Picwing before, which was also a digital photo frame type company. And it's funny that he was on that before. It wasn't particularly successful. And then boom, he hits Gusto and it's a massive company, becomes a unicorn. So I think behind a lot of these unicorn founders, they actually have quite a lot of interesting medium or low success type companies behind it. And then go on to do something bigger. Like Howie, he had eTax, sold to Salesforce for some amount of money, but then Airtable comes afterwards. So yeah, there's some really interesting repeat founder stories, I think, dotted through the Valley.
Yeah, it definitely seems to me that if I had to pick a trait out of really successful entrepreneurs, persistence and general hustle would probably be Top of the list. I know part of your ability to have such deal flow and find these angel investments has been you've attended, correct me if I'm wrong, but every YC Demo Day since you were a part of the program.
Yeah.
And you're taking a lot of notes. I was reading the intro to your rolling fund and you mentioned that you take a lot of detailed notes on why you would invest, why you wouldn't invest. What are the common things that you're coming back to when you're taking notes on these presentations? What are the first things you're looking for whether to invest or not to?
I mean, the presentations originally were so variable because there wasn't the training and the standardization around them that you could just tell straight away whether something's going to flunk or not based on like the presentation quality. And it's funny, even like Paul Graham was in the room and he'd be standing there on the right-hand side, hand against the pillar, and he would just leave the room being completely pissed off if someone was messing up the pitch, including on demo day. And he would just leave the room almost out of embarrassment or being pissed off or something. He would just walk out. And I was like, whoa, okay. So that's a good signal. And, you know, did he walk out or— but you could also just tell directly. But then over time, the presentations became more refined, a lot more training. So it started to hide itself in the details of other things like the model, the margins, their go-to-market strategy, their traction, their pre— do they have LOIs? Is the market large? All the kind of fundamentals that really mattered. So trying to take notes on all these different things, different kinds of signals, obvious, non-obvious, things that gut feeling, first principles, and like, do my fantasy investing in a sense, right? And then I started to be able to put money behind it. And I was checking back against my notes of like, did that work out to be a good company? Because I'd had different scoring systems as well. And Imad had his scoring systems, other people had his scoring system, had his or her scoring systems. And we were discussing scoring systems even, like, should you do 3 ticks, 1 tick, 2 ticks? All this kind of stuff was happening in the audience and being traded, like traded for free, you know, just everybody sharing this information. And people were starting to become like really decent investors just by seeing as much data in front of us, right? And getting to see the outcomes. And yeah, that was a pretty good bootcamp for understanding companies of a high variety. All different types of fields. Then started extending that outside of YC channels and looking at the network that I was involved in and who's making companies. Here I am now, it's like 200+ investments later, 10 unicorns from early stage, many first checks in really cool companies. And still at the beginning of even understanding how this stuff works. And it's changing all the time. That's the best part. Crypto has occurred and there's completely new markets there and biology has expanded and become huge and important. And everybody's become an investor in some sense. We've seen the rise of rolling funds and solo GPs. And that means there's a lot more companies around the world and there's a lot more investors as well. But I think this is all a good thing. There's just more people involved in innovation now.
Yeah. Well, it's amazing. You mentioned 10 unicorns already, and when I look at your track record, it's over the course of a decade. A lot of these companies take several years to play out, and so I imagine you've planted, like you said, over 200 seeds, and some of them have already grown into big successful trees, but a number of others are probably— you might have some ideas about which ones they may be, but they're probably baking to to be pretty successful companies, especially in some of these newer spaces that you mentioned like crypto, bio, whatever it might be. You don't have to get too deep into it and I don't want you to reveal the secret sauce or anything like that, but when you're talking about this scorecard that you're keeping and you're trading notes with other people, loosely, what does that look like? Are you giving a couple points to 1 to 3 to the founder, to the idea, to the market, et cetera, or how do you generally— approach doing such volume?
I mean, part of the approach is actually focusing on the ones you don't want to do and writing them off, writing it off quickly. And you're left with ones that you can't find reasons why you wouldn't do it. So I don't— there are so many different ways to look at the startups. You could look at the founders, the founder chemistry, the psychology, and different elements in there. And it depends on their stage. So if it's earlier stage, it's going to be more founder psychology, founder chemistry, the lead founder, what's behind them. And if it's further along, it might be the market they're in and the product they've built, product quality. And there's so many variables that you can tap into. You are looking for some kind of key thing that's saying, yes, this is the one to go for, or no, this is the one to not go for. So I don't really do it in a kind of like weighted scoring type system. It's just like the resulting thing became a score, but that's gone away. I kept changing the model over and over and over again. And now it feels quite intuitive now. And you can come up with like, you can do a check. So I had all these checks that I was doing, like business model, market size, founders, founder chemistry, how interesting is it, can I add value, all these prescriptive whitelist checks. And blacklist type, oh, that's a bad thing. They have a broken model here, or the founder team looks like it's going to fall apart, or they've got 4 founders or 5 founders. You can build these filters up and down, and eventually it becomes fairly intuitive to say, is that an iconic company? It's relatively spaced, so we're going to do 3D print a rocket with Tim, given his background, and they built the largest 3D printer in the States that can do metal. Is that an iconic company? I think it is. And now it's a unicorn. So you can come up with a thesis, and that could be in different areas. But I guess over time, it starts to become a little bit more intuitive, and you don't even necessarily have to say— you don't have to break it down necessarily. But I'm still learning here. And I think as the market changes, I don't want to get Things are changing all the time, and this window of opportunity of angel investing, how it's working, the model might completely shift in the future, right? We saw a little bit of the start of that with ICOs, but then it got kind of clamped down. And that was about to upend all of that type of stuff, because it was going to be very different. So I think— Yeah, I guess you've got to find your own path by— you've got to find your own learning, recursive learning algorithm where you're going to be testing a hypothesis. The feedback cycles are slow, as you said, right? It could take 7 years or even a lot longer to find out if it's going to work out or not. A lot of the great companies actually, you know, straight away, you know, very early on as well, if it's doing really well. And some of them are late developers where no one believes in it for 3, 4, 5 years, and then boom, they start to get there. So yeah, I guess it's very messy. But yeah, I think there's— if you really care about it and you're willing to spend time learning it, then you can get there.
Last question, and then I want to go ahead and talk about Golden. But you've been following crypto. You mentioned early conversations with Albert and Naval and Chris back in the Haze App days. You've been following it for a while, I imagine, and now several years into it, like a decade if we're being technical since the launch of Bitcoin. What has been— what has the ride been like watching crypto develop and what do you foresee for the future?
I think it's a really interesting area of a completely new landscape of what companies can be built. Like all the classic companies can be re-engineered, like to be decentralized possibly. I mean, the mechanisms of financing, the mechanisms of incentives, of communities, of systems, of networks, the decentralized nature of it, the disruption on the classic— let's take like Unstoppable Domains, for example, you know, or Urbit and Kind of new ways to disrupt older, more established businesses where you think you could never disrupt the domain industry or GoDaddy or some of these older players. I think that's really interesting that everything is possibly remakable in a new way. And then it's all these things that help the system work. and the ecosystems expand. So for me, it seems to have unlimited potential. And I'm very excited about that. And I'm not even going to call myself— I'm not an expert at all. I'm just an absolute beginner, even though I've been trying to keep tabs on it since pretty early. And taking part in Ethereum presale and buying Bitcoin around $30 or so and being an enthusiast and trying to get watch different projects and learn about different consensus mechanisms. And yeah, so it's an area that you could probably dedicate all your time to, to get good at. I'm also interested in other areas like biology, space hardware. So I tend to dabble in different areas. But yeah, it's something where I'm a non-expert, but I'm always shocked and inspired by some new projects that people talk about. And the game is not over there at all. The innovation roadmap could be pretty large.
Yeah. And certainly interesting. You seem to have a super wide area, a large spectrum of different interests from, like you mentioned, crypto to space and all of these different areas. And some being super technical, like I was on your website. And you have this list of ideas that you kind of are looking for people to do so that you can back them. And you have some that you've already been able to back, like Superhuman. You wanted someone to fix broken email. And then you've got a company that makes better products for— like the Dyson for baby company, I think was what you called it or something like that. And then I get to the bottom section and it's like half of the words, I don't even know what they mean.
Yeah, let's put them out there for fun. And some of them are not practical. They're just like, they even just like, you know, they may be very cool experiments, but not practical businesses. In fact, most of them looking there, just not, you know, I don't put all the practical businesses up there all the time. I just put like ones that I thought were interesting. And yeah, some people are doing it. Like, it's actually really cool that people are building these things. I was thinking of GreenCoin and like, how do we build environmentally friendly mining? And then it's we saw some cool stuff, like computationally more efficient mechanics. And the blockchain and synthetic biology, could we do something between these areas? I thought that there could be interesting applications. And mixing and matching different things, models and stuff. And things do pop up, and then I go after them. Or I get told about them, and I'm like, wow, that's new. I didn't think of that. And lots of things, everything is basically broken. To me, all the products are broken currently. Even the ones that we're building are still broken. In 5 years' time, 10 years' time, you're going to look back and say, that's pretty broken. And maybe we even go backwards as well, which is scary. Like laptops, they feel like— some Mac laptops feel like they've gone backwards. And some software feels like it's gone backwards as well. So yeah, I think this is good in that there's loads of opportunities for new people to come into the ecosystem and build new startups, new products. Everything is open to do. It's even better nowadays than it was before in terms of getting funding and getting talent and the playbook being a bit more known. And it seems like there's infinite numbers of things we could go build, which is pretty exciting.
I feel bad we're just getting to your current company, Gholdan. It certainly deserves a lot more time to talk about, but let's just introduce it if you don't mind. You're trying to map out the world of human knowledge. What is Gholdan and how are you approaching that mission? Yeah.
So over the last 10 years and looking at a lot of these companies and technologies and thinking about just doing research and thinking about how things work, I kept running into this data issue of not being able to find the information. And, you know, information spread over the web. It's on the website, it's sometimes on Wikipedia or Crunchbase or PitchBook, or it's in this document over here, or it's in a presentation over there, or an archive paper, archive paper, or a patent. And it's just all like a database, open database over here. And it's all the information is so spread around. So, you know, what we're trying to do is build a knowledge graph, large-scale knowledge graph of all the entities that exist. And let's say 10 billion things, all companies, the products associated with them, all the technologies, the people, and put it all together into a queryable database so you can do things with it, both on the app front. So maybe I want to build a list of something, a list of regenerative medicine companies, or I want to have alerts if there's any quantum cryptography companies that are coming out that are in France. Or I want my API— I want an API feed to populate spinouts of these types of companies or experts in hydrogen fuel cells. So this stuff should, you know, if you ask yourself, there should be a database of knowledge that humans know about. And there are some different types, different previous attempts, but it's not very— it's not super comprehensive. And, you know, that's what we're going for. We want to use modern NLP, like, you know, entity extraction and event detection and entity linking and all the different, you know, NER and natural language generation to try and do this and accelerate and automate the process. Initially, we're going after companies, focused innovative companies as well, because that's a pretty cool area to go focus in and build a core and the technologies that surround it and the people related to them as well. And so we got backed by Andreessen Horowitz, Founders Fund, GigaFund, and then we got backed by Andreessen Horowitz again for our A. Marc Andreessen joined the board. We raised about $20 million. We now have paying customers as well for the commercial version. And it's really cool. You can go to golden.com and look up the topics. We don't have everything yet. I mean, feel free to add your startup or your entity there. Make sure it's canonical in its writing and neutral. And if we get it right, then over time, then we should have this large-scale knowledge database, which doesn't exist, of all the things out there. And that should be really useful for improving everyone's education and helping companies as well. And I do stress that the pages are open. The individual pages are open as well for people to go read about what is this entity, what is Neuralink, what is GC Therapeutics, what is Graph. network theory? What is benchside? What is Illumina? What is a proof of proof of space? Proof of proof of work? What are the different kinds of qubits? These are the kind of things that are really interesting. And it's not fully served by Wikipedia as well, because of the notability requirement of topics on Wikipedia, and because of the schema and the taxonomy not covering everything. And then certain functions that we want to be able to do that are missing there. And Wikipedia is awesome. You know, we spend— I spend a lot of my time on there, and I know a lot of other people do. But, you know, it's 19 years old now. It's time for— it's time for something new. And, you know, we don't want to forget what they did, what they've done, and build on the work. And, you know, these pages, we've opened them on Creative Commons 4.0. So this does have, like, more than corporate interest and is actually useful for an individual in the middle of a village, someone that's trying to learn about synthetic biology. And doesn't have the money for a textbook. But we want to see more open information.
Yeah, it's cool. I spent some time on the website and looked at a couple of topics and then also was able to— there's these AI suggestions that are almost comparable in my mind to spell check, where it tells you what to— spell check automatically goes through your document and tells you what to do. And this is basically like making suggestions. And as the human reviewing it, you get to tell it, you know, that seems right or that seems not right. How critical is that human kind of AI connection to building out this system of knowledge that over time should be more comprehensive, I think, than Wikipedia is the hope and much more organized than something like Google?
Yeah. So Google is like super important first pass on the web, right? scan everything and know where to point the user to. Like, hey, go to this website. Hey, go to this website. And then if you look at it as Google Knowledge Graph and it fulfills some of the kind of functions of like, hey, this is the canonical answer to this entity question or whatever. But it doesn't really get what we really need now. And I'm saying, well, why don't we just— we need to get to the data. The data's throughout all these different public documents. So it's not good enough to just say, hey, go to this area. I want to say, this is the answer. And I don't want the answer to be fuzzy. I want the answer to be the right answer. So that's in respect to search. Search is getting a bit stale as well. If you think about it, Bing, Google, DuckDuckGo, they all have pretty similar interfaces and output and they have reasonably similar answers. They throw ads in front of it, which is And sometimes the correct answer, sometimes it's a massive distraction. You're looking for company X, they're showing you the competitor, which might be helpful to some people. So I do think there's something there where, could we come back to something much more precise where you're really looking for X and you get X? Yeah. And that, I think like 20% of queries out there result in some entity or a collection of entities or a question about an entity. And we want to be the answer for that. And then everything's shifting to voice. The problem area is shifting to voice search. But to me, the answers are the problem. You can ask the question and half the problem is turning that voice query into database lookup. But then you've got to get the answer. And that's the bit that we want to really focus on, of being able to actually answer it and having the data that would be able to explain it. But this is a very hard problem. So I don't expect this to be easy. We would be very lucky if we managed to pull this off or even crack a subset of the problem.
So, I mean, obviously you're talking about how difficult a problem is and the challenge that lies ahead. I know you've spent up to a decade with ideas that you have on the back burner, waiting for the right time and the right company to invest in. Obviously, you're running this, not necessarily investing in it. It's much more hands-on. You're committing your everyday to it. Why is now the right time for Golgon?
Yeah, so, well, I think if you look at natural language generation, if you look at last year, this year, this is the real golden age for that. You've seen all the interesting things happening with GPT. And if you think about the collection of off-the-shelf algorithms now in terms of natural language extraction, like entity recognition and linking and taxonomic detection and cluster detection, Sorry, clustering of entities and detection of events and all the problems out there. There seem to be enough of the pieces to try and put something together in terms of the technology. Compute as well is getting cheaper and more commodity in how to apply it. And then we've learned on the app side, we've got all this other stuff that models that are decent, like GitHub. And it may be that we're at the right time where there's enough pieces of the puzzle to put together something a bit more off the shelf. I think about it as well in terms of boom, this is the right time for supersonic jets. And Concorde did it at a certain point in time and now we're back and there's things that have moved along in terms of the technology. So, you know, I hope the timing's right. And it certainly gets better all the time. Like, you know, the timing gets better all the time. And then, you know, we think it's right timing for this. And, you know, we certainly can solve like problems at the easier side of the problems to let the time ticker go along a bit further to get into the harder problems. Like auto-claim validation and natural language generation, canonical natural language generation where it's all correct information, no fictitious statements. So yeah, we will see. And this is the big question, right? Because if you get the timing wrong on something, then you're in big trouble if you're too early. And if you're too late, well, you better be the best because your competition would have already built a network effect. So let's see. And I think there's things you can do to be clever about that. You can solve certain problems that monetize and grow the network effects of the business and grow your data Which I hope we're doing. I think we're doing that. And yeah.
Well, that's a good answer, I think. And it seems like you're being very honest and thinking about it. There's no way to know if it's the right time, but certainly can hope and do everything you can to make it so. I could have interjected all along the way with congratulations. It's amazing all the things you've done along the way from your exit. to all these investments, to now working on Golden. So I want to end it that way and just say congratulations on all your success today. And I'm looking forward to seeing what you can do with Golden, and hopefully you guys are able to succeed towards the vision that you're aspiring towards. I want to give you the last word to point people where they can go to find Golden or to learn more about you or the company. But thank you very much, Jude, for taking the time today. It was really awesome talking with you.
Yeah, this is fun to go into the depth here and into the entrepreneurial mindset somewhat. Yeah, in terms of pointing people, if you want to play around with Golden, if you just want to learn things, go to golden.com and then click on— try to get to explore and read through. If you need to represent your company there or your person or technology, go for it. You can add it in and follow the writing policy. If you're a commercial company, and you want to use our tools and you're a large-scale company, get in touch. We have enterprise solution for solving problems and being the brain of your organization. If you're an LP, back my rolling fund. You can find that on AngelList. If you're a startup that wants to be backed and you've built something really cool that resonates with the portfolio I have, get in touch directly as well.