The following interview is a conversation we had with Paolo Fragomeni, CEO of Socket Supply, on our podcast Category Visionaries. You can view the full episode here: Paolo Fragomeni, CEO of Socket Supply: $3.5 Million Raised to Build the Future of P2P Computing
Paolo Fragomeni
It’s not easy. It’s Italian. Sicilian. It’s one of those long Sicilian names.
Brett
Nice. Well, glad to hear I did. Okay, now let’s start with a quick summary of who you are and a bit about your background.
Paolo Fragomeni
Yeah, sure. Let’s see. In 2010, I left MIT to found my first startup, which was a platform as a service called Nojitsu. And since then, I’ve had some interesting successes and failures. But I’ve been working on infrastructure and developer tools for over a decade, and I’ve also been involved in peer to peer communities for even longer, more than 15 years.
Brett
Nice. Very cool. And a couple of questions that we like to ask just to better understand what makes you tick as a founder and as a leader. Is there a CEO that you’re studying the most right now? And if so, who are they and what are you learning from them?
Paolo Fragomeni
Well, I try to avoid that kind of thing. I try to listen to people objectively and decide if what they’re doing and saying makes sense. But at the end of the day, I think that founders are just contributors and founders anything without a team. So if I can say teams maybe are who I would look up to, I think that Tailscale is a really amazing team. Oxide Computer is a really very cool team. These are remarkable teams. I definitely get inspired by their work. They’re solving super hard, really meaningful problems that drive the field of computing forward.
Brett
And are there any specific takeaways that you’ve learned from these teams that you’re applying with Socket Supply?
Paolo Fragomeni
Definitely. I think cultural values, I think technical values, I think research that they’ve contributed to the field, we’ve definitely been able to leverage the corpus of research that has gone into the field. I think I would say yeah.
Brett
Nice. Very cool. And what about books? Is there a specific book that’s had the greatest impact on you as a founder? And this can be a business book or it can be a personal book as well.
Paolo Fragomeni
Probably The Improvised Munitions Handbook. Or I would say maybe The Anarchist Cookbook. I remember I found the Improvised Munitions Handbook at a garage sale, and it was vintage. It was from the 1950s. And then later I traded it to a street punk for The Anarchist Cookbook and these things are cute compared to what’s online these days. But as a young person I found them really fascinating and inspiring because really they made me kind of realize how anyone can learn and teach and build anything nice.
Brett
Those are two books I’ve not heard of. It’s refreshing to have something other than the hard thing about Hard Thing or Peter T. That’s a great read.
Paolo Fragomeni
All of those things. Yeah, I mean these are all really good books but I mean, we’ve all read them.
Brett
Yeah, for sure. Yeah, it’s nice to hear something new. Now let’s talk a bit more about socket supply. So what’s the origin story behind the company and what’s the high level pitch? What are customers paying you to do for them?
Paolo Fragomeni
So our AHA moment came I guess, when we started seeing these web three products and we said, wow, these are really bad, they’re not decentralized, they’re SaaS products and they’re investor subsidized EC two instances running on AWS. And this is a huge problem because it forces most of these products into negative pricing and we thought, okay, people want a decentralized web and there’s a demand for it and they also want to be liberated from sort of cost of the cloud and so we decided we would try to make that happen. So we told some investors that we thought there was a really significant technology gap in this entire space and they agreed and we raised a little over 3.5 million and we got to work for a little over a year. We went heads down and we coded. We all come from peer to peer background and we’ve been working on distributed systems for a very long time and so this was really natural for us.
Paolo Fragomeni
And so here we are today with a whole open source runtime that allows the average web developer to write apps that connect directly to each other. The way that it works is developers use HTML and CSS and JavaScript, all the things that they know and love to write cross platform native applications that run on desktop or mobile. So you write the code once and you literally run it anywhere. Yeah. And so then we expose peer to peer APIs so that apps can connect directly to each other. And this is all built on UDP and Bluetooth. So local first is obviously a part of this agenda as well. A year and a half later, here we are with something that we’re going to market with and something that you can actually truly build decentralized applications with. Something that is remarkable in the way that we’ve always been sort of in this landlord tenant relationship with the cloud and that’s really prohibitive for a ton of different reasons.
Paolo Fragomeni
Specifically the cost and the complexity and the knowledge that it necessitates in order to take anything to market makes a lot of sense.
Brett
And what types of developers are you seeing really adopt the product and embrace the product? Is it a certain vertical or a certain company size? Or are these just independent or freelancers? What does that look like?
Paolo Fragomeni
Well, we’re seeing, obviously, public sector startups founders, creatives, basically anybody who wants to build anything that they would normally build for the web. They’re realizing that they can build it with the socket runtime but not pay any rent to anyone. So, yeah, anyone who builds web apps or software, native apps, they’re all interested in what we’re doing. But recently we’ve been talking to the folks in the German government. There are certain things I can’t say about it, but the other things that I can say is that cloud vendors are expensive. And there’s this due diligence process that’s really sophisticated and lengthy. It take a note of time. In the US, it’s known as FedRAMP, and most countries they have something like this. It’s expensive, it takes a lot of time, and it could be prohibitive to shipping anything. So now imagine that you don’t need the cloud at all.
Paolo Fragomeni
And imagine your it controls everything. Suddenly you have autonomy and there’s a smaller surface area and less potential for incidents.
Brett
Interesting. I feel like that’s becoming more of an active conversation. Right. Like the move away from the cloud. I can’t remember what’s the parent company of Base Camp is. It 37 signals.
Paolo Fragomeni
Yeah.
Brett
Their CEO just had a blog post of why they’re leaving the cloud. Did you see that? Yeah.
Paolo Fragomeni
And it’s really understandable. It’s an insane cost. And it’s not just the cost of like well, what does it cost monthly? Right. Because for some people that it’s a trivial cost. Especially with free tier, it can be pretty inexpensive. But once you actually start to get to the point where you actually have demand, then it becomes nontrivial cost. In fact, it becomes a cost that, again, can be prohibitive.
Brett
And what are you doing to convince customers, convince developers to move away from AWS? I see that on your website. Transition from AWS to P. That has to be a hard competitor to go against. Right.
Paolo Fragomeni
AWS is very established.
Brett
They have a lot of resources. So what are you saying to them and what are you really communicating to them that’s making them willing to make that kind of jump?
Paolo Fragomeni
Well, I think the timing is everything, especially in a downturn when everybody’s stock is in the toilet. We have these very sage C suite executives asking where’s the burn and cloud is. A huge burn and supplementing cloud infrastructure with peer to peer can bring your cloud costs to zero. And this is just a foot in the door. When the cloud was created, there was a fraction of the hardware in the world.
Brett
Right.
Paolo Fragomeni
AWS made tons of sense where they were there to provide high availability. And today they’re doing more the same. Right. They have roughly 38 data centers and I don’t know how many availabilities, like 85 or 90 availability zones, 27 regions, and about, I think maybe 50 to 80,000 servers in each data center. So roughly like 3 million servers. But all that pales in comparison to the compute and the storage that’s in the wild, right there’s about 15 billion mobile devices today, three or five something billion laptops, and just an enormous amount of IoT devices and sensors are in everything. There’s so much hardware in the world right now, and the more hardware there is, the more it can be connected and the more it will be connected and the less it’ll be making these round trips back to the data center. These are expensive and they’re really super complicated.
Paolo Fragomeni
And so all of this hardware is forming these like ad hoc peer to peer networks and the cloud is really going to be sort of more used for supercomputing tasks and all this stuff’s really going to move away from the cloud. It’s something that really is already happening and we’re just here putting tools onto it so people can leverage that fact. Okay, so when I say this, there’s some people who it’s intuitive, they get it. They’re like, I totally get that. Then there’s some nerds who have sweaty palms and they’re starting to get really angry because they’re saying like, oh, you can’t compare a smartphone or a laptop with an AWS server. And of course you can’t. None of the hardware in the wild is a one to one replacement for a server. If you try to monopolize a user’s resources and treat it like a server, the user is just going to delete your app.
Paolo Fragomeni
But it’s unnecessary to try to do that even because even at a small scale, small contributions or bursts of being online or storage and compute, all of this adds up when there’s this much hardware out in the wild. So imagine everybody scrolling instagram was forming a peer to peer CDN. You could cut bandwidth and storage and caching all those kinds of costs. You could cut all of that out of your burn. And imagine people all scrolling Twitter at the same time doing something similar. Now imagine if you could piggyback these networks could piggyback off of each other. So now your average web developer can do this. And that’s what we’ve created. And that’s the problem that we’ve solved, is like making peer to peer available to the average web developer so that you can escape the complexity and the cost of the cloud. And I think a lot of people see this really clearly, really quickly and it becomes intuitive.
Brett
And what are you doing to really stand out and cut through the noise? Because I feel like P to P computing got a lot of hype, obviously with crypto and blockchain, all of that stuff. So how are you separating from all of that noise and all of that buzz that’s been created by that I.
Paolo Fragomeni
Don’T really notice the noise. I don’t really care when people are talking about what the hype cycles look like. I mean, I hear about things, but I think when you have something that clearly solves a problem and it has a very clear market and that market has a demand for it, then there’s not much of a conversation other than is your team capable of shipping that solution? So we raised, like I said, like three and a half, about a year and a half ago. We have really low burn, we have about 30 months of runway left. So we haven’t really had an urgency for conversations about raising. We still take calls from people and we still have these conversations. We have about like 6 million soft circled for follow on for a seed. Like we’ll probably raise a seed, but like I said, it hasn’t been hard for us to do that because I think that what we’re doing, especially now that we’ve gone past the precede and we have something tangible that people are actually building with.
Brett
And has it been tempting for you at all to go out and try to raise fu money like a lot of other startups and a lot of other founders have done over the last couple of years? Because it sounds like you’re taking a very disciplined approach to building with 30 months of burn. That’s a lot. And that has to feel good. And I’m sure you feel hopefully you can sleep better at night than some of those other companies out there. But was that tempting for you to go down that other path or did you always know this is how you wanted to build the company?
Paolo Fragomeni
Well, I mean, with interest rates at zero, it was a party. People were doing crazy things, there was insane valuations. It was very easy and very tempting yeah, to take money. But I mean, there’s consequences to taking money and I think that you need to have a very good reason and really specific plan for taking that money. So I mean, I’ve seen people get into a lot of trouble taking more than they needed to do what they’re going to do, right? And we also see a path to profitability in the very near future. So we’re really looking at it like, only take what you need. And we started to see really good traction with people building on what we have so far. For instance, MetaMask is in the process of rewriting their application using our platform, the Socket runtime. We’ve had discussions with Bloomberg about how Belo did Electron is and the potential of replacing it signals desktop team.
Paolo Fragomeni
We’re talking to these guys about they’re actively evaluating what we’re doing. We’re talking to these different types of obviously a lot of smaller companies who are also in our portfolio. But we’ve recently just open sourced this stuff and we’re starting to see interest very quickly at a very good rate. And obviously our open source is a funnel to our paid products. So we’re happy with the rate, with the velocity, interest in what we’re doing. But it is a very new space, right. Peer to peer is a whole new class of software that most people don’t have a lot of experience building. And we’ve aimed to make it very simple and very easy for people to dive into, but it’s going to necessitate a whole new class of tools. So as a company, we’re there to provide those things and I think, yeah, we see a very short and very realistic path to profitability.
Brett
And are there any numbers you can share just in terms of growth that you’re seeing, or adoption?
Paolo Fragomeni
Yeah, I mean, they don’t really sound like impressive numbers, but like tens of new contributors who are actually really competent programmers coming in from the outside and just becoming active contributors is actually a really great metric. And I think especially when some of them have tens of thousands of GitHub followers. So GitHub followers are actually really hard to earn. A lot of companies go out and they’ll purchase Twitter followers. That’s really easy. Really? Twitter followers have almost no value in this industry anymore. But something like a GitHub follower is different because that’s an engineer saying, oh, this person’s producing something valuable. I think I’m interested in that and I’m going to follow them. They’re harder to earn. It’s harder to game that system, though they kind of represent a better metric. So if you have somebody who like we do, with tens of thousands of GitHub followers who says, oh, this is an interesting project, and they start working on it, that’s an interesting metric.
Paolo Fragomeni
I don’t like to use these types of metrics at all because we’re talking to maybe an engineer who has no avatar at all or something like that. And then you discover, oh, this is the VP of engineering for this slightly large company, and that might not be representative because then you have twelve people on their team and then turns out they’re all using the platform too, and they’re building their next version of the application with that. So it’s like there’s a lot of hidden relationships that don’t get on a visible graph. So I never like to really think about these kind of metrics. I think some of our internal metrics for gauging, do people care about what we’re doing and how is it being received? Is some of the people who have been in the space for a very long time taking interest in actively picking up the product and trying to use it?
Paolo Fragomeni
Right? We have a few folks who are senior staff engineers at Uber who are interested and they’ve joined in our channel and they’re talking to us and they’re experimenting with it, I think, things like that. Those are hard to describe metrics, but they’re really significant, I think, and that’s really good validation for us. Internally.
Brett
Nice. I love it. Last question here for you. If we zoom out into the future, what’s the three year vision for the company?
Paolo Fragomeni
Well, I see our tools and I see our open source having significantly displaced cloud services and the things that are built on them. So all these platforms service, all these SaaS products where you paid rent at one point on a cloud service, you’ll be spending that money developing features on your app instead, and you’ll sort of be free of vendor lock in. And you have regained this autonomy that you lost being a tenant, paying rent to the cloud, happily using our tools.
Brett
Amazing. I love it. Unfortunately, that’s all we’re going to have time to cover for today. So, before we wrap, if people want to follow along with your journey as you and the team build, where’s the best place for them to go?
Paolo Fragomeni
They can join into a whole bunch of different places. So at the footer on our website, there’s a bunch of different options. There’s GitHub, there’s Discord, and in Discord, people can ask to join. We have a fully peer to peer, completely decentralized, absolutely uses, no servers, chat application, which you can join in or obviously Twitter. There’s a whole host of whatever works for you.
Brett
Amazing. Well, thanks a lot for your time. I really appreciate it. It was really fun hearing about what you’re building and wish you best of luck in executing on this vision.
Paolo Fragomeni
Yeah, good talking with you. We’ll talk soon.
Brett
Take care, you.
Twitter
facebook
linkedin