First things first: the “Hashi” in HashiCorp is pronounced haw-shee. Not hash-ee. But don’t worry, most people get it wrong, and Mitchell Hashimoto isn’t about to correct you. He and his co-founder Armon Dadgar would mostly just marvel at the fact that they’ve built a company large enough to be so widely mispronounced.
The privately held, 625-person company, started in 2012 by these two University of Washington computer science graduates, has emerged as one of the new powerhouses of the cloud, developing a large and loyal following by focusing on technologies to deploy, secure, connect and run infrastructure as code.
Based in San Francisco, HashiCorp is part of a new wave of new companies applying modern software development practices to technology infrastructure — the plumbing of the cloud — putting a new spin on what was traditionally a manual process of provisioning, configuring, securing and updating the underlying hardware and software that runs modern applications.
It’s a big market. HashiCorp entered the ranks of the industry’s unicorns last year with a $100 million venture capital round that valued the company at $1.9 billion.
— HashiCorp (@HashiCorp) September 13, 2019
Dadgar and Hashimoto were back in their old stomping grounds last week, along with HashiCorp CEO Dave McJannet and other members of their team, for the annual HashiConf community conference. The event drew an estimated 1,600 attendees to downtown Seattle.
In a conversation with GeekWire behind-the-scenes at the event, Dadgar and Hashimoto reflected on their startup journey so far, opined on the current state and future of cloud infrastructure, told the story of how their headquarters ended up in San Francisco, and explained why the location of the HQ actually doesn’t matter as much as the reporter talking with them might think.
Continue reading for edited excerpts from our conversation.
Todd Bishop: What does it mean to you to come back here to the Pacific Northwest, where you got started?
Mitchell Hashimoto: We started the projects and never planned on making a company. It’s totally fine that some people have ideas and they do it to start a business, and that’s normal. I think more normal. We were just solving our own problem, very committed to the jobs we had at the time and solving their problems, and so the fact that it’s come this far is just shocking. And when you’re in it, it’s hard to step back and take a look. When you do things like fly over UW, it’s like, whoa. We started it in the basement of that building … and now we’re back, which is cool.
Armon Dadgar: My parents still live here. I grew up in this area and they actually came to the conference. They watched the keynote and they were like, “Wow, we didn’t realize HashiCorp was this big. You’re on stage with Microsoft and VMware here.” They’re like, ‘We didn’t even realize these people partner with the company.” It’s funny because we’re so in it that you kind of forget. You don’t get that chance to step back. When we started, we thought we’d we spend six months working in my living room and then, you know, we’d fail and go work at Google or something.
Hashimoto: It’s a nice living room, though.
Dadgar: We had a little startup group where we got five, six people together and we cycled through ideas all the time. And it’s funny to just think about the ambitions then and expectations you have. And you’re like, well, what we would’ve considered success then would’ve been growing to like 10 people and having like five users. So it’s kinda cool to be back here and reminded of that.
TB: You like to say that you love infrastructure.
Hashimoto: Yeah. Weird, right?
TB: What do you love about infrastructure?
Hashimoto: I’ve always loved the automation side of things. [Infrastructure] is a place where you could automate and have an immense amount of impact, on whatever’s running above it, which, since everything’s technical, it’s a huge, broad scope. I got into computers because I got a kick out of writing some code and seeing a computer do something without me doing anything at the time. And so I’ve always automated things and I think as I started getting into software engineering, both academically, my first jobs and so on, I just saw various manual processes which I thought it’d be cool to automate. Can I automate that? And that’s been the motivation for all our tools, more or less.
TB: When you look at the cloud and infrastructure automation in particular, where do you feel the industry and your company are, in terms of the long-term evolution. The analogy a lot of people would use would be, what inning are you in? But baseball’s not always the best comparison.
Hashimoto: I grew up playing baseball, so that’s fine with me.
TB: What inning is cloud infrastructure automation in right now?
Hashimoto: Well, it’s early, unfortunately. Cloud infrastructure automation, I guess, is doing all right. I think that cloud infrastructure automation’s in maybe the third or fourth inning probably. I think overall infrastructure across the board is still super early. One of the big challenges is that we’ve had a lot of new over the past five to 10 years, closer to 10 than five. We’ve just had a lot of new and there hasn’t really been that period of stability. People are getting value out of everything, but there hasn’t been that period where we’re comfortable with everything that we have, and businesses are just pulling value full steam out of these things. They’re still spending a lot of money to adopt cloud and migrate legacy infrastructure and so on. There’s so much migration into cloud still, and so much legacy. We have to get to that period of stability. I think our products have to get better to get there. Other products have to get better to get there, but we’re getting closer than ever before. The fact that the enterprises are now full swing into the cloud adoption shows that we’re heading towards that.
— Jay Fry (@jayfry3) September 11, 2019
TB: It struck me that you said, “unfortunately,” because another way to look at that is, lots of runway, lots of space to grow. When you look at where cloud infrastructure automation is right now, what is the potential of this company?
Hashimoto: The potential is huge. We are deep in three different industries, at least, which is security, operations, and networking. If you look at the various big vendors in each of those categories, you could name them, we’re partners with a lot of these people, so I’m not naming these as competitors, but if you look at VMs and you look at VMware or you look at what Hyper V has done or Xen; and then you go to networking and you see like Cisco and Palo Alto Networks and Juniper and those things; and then you go to security and you see CyberArk and Thales and all this; we have a good opportunity to own a lot of those markets in a cloud-native world, and so I think the opportunity is huge.
TB: When you look at the big cloud providers, people are deploying to those cloud providers using your tools, but at the same time, many of those cloud providers — AWS, Azure, Google Cloud — have tools that could replace yours. Do you view the big cloud providers as competitors or partners?
Hashimoto: Partners, for sure. They have competing-in-functionality products for their single clouds. But we do it across clouds and we have a lot more tooling around that. And when customers approach cloud, they’re looking for an investment, a strategic investment that they could make in a vendor so that they’re future-proofed as much as possible, so that they could get the governance and risk mitigation and all that other tooling that they need. And we offer all that, and the clouds don’t mind because we don’t hide their platform behind anything and they don’t charge for these tools. The clouds only have these tools as a way to get you to use them.
Dadgar: The analogy I like to use is, they’re in the business of selling you power. Fundamentally, they don’t really care what you’re doing with the power. We’re in the business of selling people appliances that let them draw more power. And so from the cloud providers’ perspective, there’s no conflict. We’re not the competing power vendor. We’re saying, “Hey, we’re going to help your users draw more watts, faster.” And the cloud providers are like, great. So I think we’re economically very aligned. We sell tools to let people use the cloud more easily. As you saw, folks like Amazon, Microsoft, Google, VMware are all platinum sponsors of the event.
TB: Who then are your biggest competitors, in your view?
Dadgar: There’s a lot of the legacy vendors who have a more traditional approach to infrastructure management. That’s one side of it. But I think, in some sense, it’s a new market. So it’s not like there’s this stagnant market and the market’s fully divided and we have to eat market share away from other existing incumbents. There’s an existing on-premise world that’s been carved up and there’s the incumbents who own it. Fine. Set that aside. There’s this brand new cloud world that’s all greenfield, basically, where there’s not an existing incumbent you’re displacing. These are net new workloads. These are net new applications and services that people are building.
TB: Would you have ever imagined that your company would be this size? What’s been the secret? Is it luck, hard work, recognizing the market?
Dadgar: There was a confluence of a few things. I think the timing was critical. You had to have the right partner ecosystem around you.
Hashimoto: And I think we got lucky, actually. Being in Seattle at UW was really lucky. Amazon, Google and Microsoft are here. Microsoft donated something called Data Center in a Box. And then Amazon had EC2, so they gave us credits. Google gave us servers. Universities are really good at using as many of those donated resources as possible. And so that’s where a lot of this multi-cloud thinking and workflow agnosticism came from. We had to make those worthwhile, because they were the only resources we had.
TB: You’re both here in Seattle once a quarter or so. Two of the major cloud platforms are here. And the third is almost here at the new Google campus. And in some ways you can look at HashiCorp and say, “Oh, it’s the ultimate Seattle startup success story,” and yet it’s not a Seattle startup success. Talk me through the process that ended up with the headquarters in San Francisco instead of here.
Hashimoto: We’re a deeply technical company and all of your initial adopters and users, for better or worse, the deeply technical companies, are also in the Bay Area, particularly five years or seven years ago. So while there are technical companies everywhere, there’s just such a concentration there that when you create tech for tech, it’s very beneficial to be present in San Francisco. I moved to LA six months after we founded the company, so even though we headquartered it there, I moved almost right away, but it was still so important. I was flying back and it was just so important to be able to get an email from somebody and they say, “Hey, we’re using your product.” And then Google it and be like, “OK, well, you’re next door, so could I just come over?” and that feedback cycle is really important early on.
Dadgar: It was a bit of happy accident. We ended up after school working at the same startup together. It was called kiip. It was a mobile advertising company that happened to be based in the Bay Area. And so it was sort of accidental. When we left kiip to start HashiCorp, we happened to be there. And so the headquarters just sort of got built there. I think that network effect is not only just that the companies were helpful, but the capital was there. I think it’s changed a lot in Seattle in the last seven years.
TB: A lot of companies from the Bay Area have established engineering centers here in Seattle. Is that on your radar?
Dadgar: What’s funny is we don’t really have centers because we’re such a remote company. The only office we have is San Francisco and the rest of it is totally distributed. So we frankly probably have more engineers that live in Seattle than live in the Bay Area because of the distributed nature of it. The office is actually not a engineering hub. It’s more of a [general and administrative] hub more than anything. It’s finance, legal, HR, sales management. But there’s maybe two engineers out of that office. What’s neat about HashiCorp is, because it’s so distributed, people are everywhere.
How to stay connected in a #remotefirst engineering culture? A lot of thought goes into practices/traditions to help collaboration/communication: lightning talks, chat roulette, weekly all hands meetings, & annual all-company meeting "HashiCorp Exchange." – @psomal #hashiconf pic.twitter.com/oLrH6Fek4I
— HashiCorp (@HashiCorp) September 11, 2019
TB: What would you say to people in the community here who would look at HashiCorp not being based here as a failure of Seattle?
Dadgar: That’s funny.
TB: You could be one of the rare unicorns of Seattle.
Dadgar: I feel like it’s improved, and some of this was feedback back to the university. I think what happens, really interestingly in the Valley, is you have this interplay between Stanford and Berkeley and then the venture community and the big tech companies. And it’s the dynamic revolving door between all of them. You don’t get that as much here. Here it’s very much, the university is almost a funnel into the big tech companies. The joke was like, there were five possible places you could go after UW CSE. There was really no interaction at all with the venture community. Outside of us caring about startups and doing our own research, there was nothing formal at the university that said, think about it, or there’s a fair about startups, or —
Hashimoto: There is now.
Dadgar: There is now, so there’s been investment, but at the time, unless you cared about it and you were reading TechCrunch and Hacker News, it was a foreign world. Like it didn’t exist. You’re going to graduate and go to either Amazon, Microsoft, Boeing, Starbucks. Those are your choices.
Hashimoto: It was crazy when we graduated, we could just ask everybody, what are you doing, where are you going? And we were the only people going to a startup. We weren’t starting a startup, just going to a startup. We were the only people. It was crazy. There was like four people. Our friends group was like the only people.
Dadgar: That investment that you now see them making, the CoMotion center at UW, and trying to bring in more of that investment and create a space, that is super valuable. I think, hey, if we had that, if there was more of a local scene, maybe instead of moving to SF to join a startup, we would have joined a local startup. And when we would have left that and started HashiCorp, we wouldn’t have left Seattle.