Peter Relan is not exactly a household name, but he has seen a lot over the course of his career and enjoyed a fair amount of success. A Stanford grad who logged time at Hewlett-Packard and Oracle before becoming the CTO of the ill-fated, bubble-era company Webvan, Relan has, in more recent years, become known in founder circles for creating three “low-volume, high-touch incubators” that have punched above their weight.

The very first class back in 2007 included founder Jason Citron who, at the time, was developing a social mobile gaming company, OpenFeint, that he subsequently sold in 2011 to the social gaming platform Gree of Japan for $104 million.

Relan, who played operating partner, owned half the company, and when Citron almost immediately decided to start a second company that he would later call Discord, he headed straight back to Relan to help build it. (Relan’s other “graduates” include the social gaming company Crowdstar, which sold to Glu Mobile in 2016, and Agawi, an early game streaming company that sold to Google in 2015.)

Relan long ago moved on from gaming and is now focused on climate change, equality and the “ethics of technology,” but he talked with us recently about his relationship with Citron. (Relan sat on Discord’s board until last year.) He also said he has “two big things brewing — one in AI and one in climate change,” so watch this space for more information. In the meantime, you can hear our conversation here or read lightly edited excerpts from it below.

TC: You launched your accelerator, YouWeb, around the same time as Y Combinator got underway, but it was structured very differently.

PR: Every year, we were like, let’s just [partner with a] few entrepreneurs here, and we’ll give them a year instead of the traditional three months, and we’ll co-create. So we don’t have hundreds of companies. Over 15 years, we’ve incubated about 30 companies.

TC: How did founders find you?

PR: It’s all referenced by other people. So, for example, I met Jason when he was 22. He was a roommate of my nephew, who had graduated from Berkeley. I was a Stanford alumnus; I asked a couple of people at Stanford, and we found the founders from there. It was very much low key. We weren’t trying to build this massive application process. We were just looking for founders who have particular types of inclinations; because I’m an engineer, I really wanted very techie, developer founders.

TC: When you say ‘we,’ who else was involved, and were you funding this incubator out of your own pocket or did you have outside investors?

PR: I started with my own funds, but then we rounded up a network for the founders, including seven or eight investors who were executives and other founders who succeeded [including at] Yahoo, Google, Oracle and Microsoft. It was like $2.5 million for that first class, and we sponsored eight founders.

TC: One was Jason, whose first company — co-founded with an EIR at YouWeb, Danielle Cassley — was OpenFeint. How fully baked was that company at the outset?

PR: Jason is a core gamer at heart, and he arrived saying, ‘I just want something in gaming.’ There was no idea, there was no particular game, there was no particular product, and actually OpenFeint, which is a social chat platform, began as a game called “Aurora Feint” that was a really beautiful game and was really well-received, but in terms of financial success, it was doing okay. And the most interesting thing about it, inside, was actually a social platform that had leaderboards and achievements and chat. So we pulled out the social platform and said, ‘Let’s [put] the game aside and let’s scale this platform.’

TC: What was your role?

PR: I was executive chairman. He was co-founder and CEO. So we worked like business partners. He was the product guy. I was the mentor. And Danielle was involved, but she moved on. So it was Jason and me pretty much. I would help with the financings and with the strategy and business development. We did a big deal with AT&T to pre-install OpenFeint on AT&T mobile phones. So there was a lot of that kind of stuff. My goal was to be helpful.

TC: Y Combinator famously takes a 7% stake for its early checks. What percentage of OpenFeint did you own?

PR: We used to do co-creation, so we would be 50/50. Obviously, it was a big hit for YouWeb [when it sold], because we were equal partners with Jason and Danielle. In our second [batch] where we worked with Discord, we [operated] much more like YC because I wanted to actually do a little less work and work across a broader number of fields, not just gaming.

TC: It says something about your relationship that Jason turned around and came back to you with Discord. It’s also interesting that a 26-year-old who’d just made a fortune would jump right back into work.

PR: I remember when he left Gree. It was late October or November of 2011, and he just sent a simple email saying ‘I’m back.’ And I said, ‘Great.’

Again, it started with a game and — we’re skipping ahead here — three years later, the game is doing okay but not great, and the chat platform features are very, very popular and, like déjà vu, he suddenly pulls out the chat platform and launches it as Discord. So [it’s] repeat [and] play but done much better this time, obviously, right? [Discord] has now got more than 300 million, 400 million users now — well over 300 million.

TC: And now its valuation is reportedly valued at $15 billion. How did its Series A round come together? I’d heard at some point that some firms — maybe General Catalyst, maybe Accel — lost out on that opportunity. Is that true?

PR: Actually, the first round of Discord was really interesting. It wasn’t even a priced round. It was really people investing in Jason, and perhaps a little bit in the fact that we were working together — sort of like, ‘Hey, here’s a couple of people who’ve been successful with something, and here’s a guy who wants to keep going and do something bigger now maybe.’ So the first round was really a bunch of VCs who had essentially decided to write these $250,000 checks. Accel was in there and General Catalyst and IDG and Time Warner.

And then the question you’re asking is how Benchmark, a completely different firm that was not [part of] those first 10 checks and was not a seed investor, ended up getting the deal. What happened was TechCrunch Disrupt. He got to the finalist thing and people really loved his presentation style and his company and [longtime Benchmark investor] Mitch Lasky noticed it and said, ‘Hey, we should talk.’ I think it was almost so fast that if I recall, it was really just, ‘Hey, is anybody else really ready to do something? If not, I have a term sheet.’