Transcript: Omry Hay | env0 | How to Build a Product for DevOps

Opening and meeting Omry Hay

Eyal David: [00:01] Hi everyone, this is Eyal. You've reached Product Builder, where you can listen to the most interesting product conversations. My journey began in 2012, when I learned what product is at my own startup. Since then I've worked in the industry, and since 2017 my company has been advising a wide range of companies. My drive is to help all of us get better as product people. And before we start, if you enjoy listening to Product Builder, please rate us five stars on Spotify. It gives me great feedback that the podcast is valuable to you, and it would make me very happy. So yalla, let's start. Omry, what's up? How are you? How was the drive here? Fine, less traffic at these hours.

Omry Hay: [00:43] How are you? Not really, but okay. Relatively, everything in life is relative. Right, we've learned that.

Eyal David: [00:50] Okay, so Omry is the CTO and co-founder of env0, right? Right. I said it correctly, count on that. Tell us, if you can, in a sentence or two, a bit about your bio, a bit about the context that brings you here. Sure, so first of all, Omry Hay, I'm 42,

Omry Hay: [01:07] married plus three, from Kiryat Ono. We studied computer science together, wow, I don't have it anymore, many, many years ago. I've been in the industry for something like 17 years. As I said, I did a computer science degree and started getting into the industry, I was a developer, a manager of all sorts of teams in all sorts of places. Always in development, though, right? Yes, always in development, until env0 I was always in development. I managed teams of all sizes, in very different technologies, I have to admit, across moves. And that's it, today I'm one of the founders and the CTO of env0, and sort of manage the Israeli site, and mainly do product and R&D. On the Israeli site. On the Israeli site. No, product and R&D in the grand scheme of the company, but yes, R&D is in Israel, that's the product.

env0 by the numbers: people, sites and funding

Eyal David: [02:05] So really, lay it out for us for a second, what does your company look like, how many employees, how long you've been around. We've been around for about five years,

Omry Hay: [02:14] to date we've raised something like 42 million dollars. In the company we're something like 50 employees, which changes every day, depending on the day. We're split, part in the United States, my partner Ohad relocated to the United States something like two and a half years ago. Where to? Where's the second base? To San Francisco, although the base itself today is in Boston. Wow. Yes, we basically built a fairly large sales department there, and we have other people in various places working remotely in the US, some in California, some in Raleigh, in Chicago and so on, even in Miami. And then the other side is in Israel, which is product, R&D, HR, marketing too. And okay, so basically it's a product for developers, something that connects strongly to your problem. Right, it's actually a product mainly for DevOps people, the persona that actually buys and uses the product is mainly DevOps people, we also have what's called Self Service, and the idea is really to also offer this solution to developers.

Eyal David: [03:26] So yes, it comes from a painful point I went through during my career. So really expand on that point, and then we'll, okay, this will be, I'll say, this will be an episode that by its nature, we'll throw around technical words, we'll try to stop where we can, but we'll keep the conversation flowing and I can ask you some questions. So really, where in your career did this catch you, that you thought of this idea, and was it you who thought of it, how was it?

What Infrastructure as Code is and what env0 does

Omry Hay: [03:52] So I'll start for a second with what we do and what the product is, and, as you said, I'll touch on a few technical points, because the product is very, very technical. So broadly, what we do is a platform for managing deployments and cloud environments on top of Infrastructure as Code. Let me expand for a second on what Infrastructure as Code is. It's basically a term that says that if you used to manage your cloud through clicks, you'd go to the AWS console and click a few times to define some server or some database, today you don't do it that way anymore. Instead you write it in code, you write I need a server of this type, a database of this type, run the code, and then you get it in your cloud service. This gives you the ability to replicate it across environments, to know exactly what's happening, who's doing it, when it was done, an audit log around it, things that used to be manual and that you couldn't do. There are several such frameworks in the world, the first of them being CloudFormation, from Amazon, which started in 2010 and is AWS-specific, meaning you can only spin up and manage environments in AWS with it. Then came Terraform. Terraform is open source from HashiCorp, it's no longer open source, we'll touch on that later, but it started as open source in 2014 from HashiCorp. Today there's also Pulumi, there's Crossplane, there are all sorts of other Infrastructure as Code frameworks, and what we do is let you manage it. What does managing it mean? In the end, all these tools, all these frameworks, are some CLI tool you run from the command line on your laptop, and we give you a lot beyond that, meaning you can say who can do what and when, how much it costs, set an approval policy, set all kinds of role-based access on what's going on, do budget management, a lot of things beyond just running and managing those resources.

Eyal David: [05:57] Wow, so basically it started, let me just organize this a bit for myself too, it started as a product that gave convenience, quality of life, basically for DevOps, you took that and added another basket of services on top of it, and today it's basically a platform of products that runs. Can you talk just about the... or we'll also touch on the business, if you want to share, but can you also talk for a second about the self-service product, or for you? Yes, so one of the things we provide is the ability for DevOps people,

Self-service: templates and policies for developers

Omry Hay: [06:27] broadly, what they often do is get tickets from R&D that they need one environment or another, or that they need some new service or something like that. So what we provide is the ability to create what we call templates in our product, and you can define what exists in that template, what it looks like, what its policy is, you can define all kinds of things that are your standard, for example tags and so on, and basically let developers run their own environments, and on top of that also define all kinds of policies. For example, you want a TTL. What's a TTL? It means I spin up this environment, but I want it to go down at night because I don't want to pay for it, because a dev environment is ephemeral. So the idea is basically to create a platform in which DevOps people define what they do or what's allowed, and basically give it to their developers so that, to their customers, to their customers, yes, exactly, so they can run their environments and develop easily without opening tickets and without being dependent on DevOps people.

Eyal David: [07:35] Okay, that's really interesting, and there are tons of points here I want to touch on, but let's take ourselves back for a second to the point we started from. So how did this meet you, and also, if you can take us afterwards a bit through, okay, how did you turn it into a product, how did it progress, you and your partner, how did it actually progress into a product, and what was the first product, what did you start with, how was it at the beginning.

The pain behind the idea: waiting for the DevOps sprint

Omry Hay: [07:57] So I'll start a bit further back. Ohad, who is my partner and the CEO, and I met at Ituran, we were sort of team leads and group managers there, we were colleagues. And we always had the problem of where are my environments, what's happening with them, who do they belong to, when can I bring them up, take them down. They were very static, meaning every time I needed some new environment, and back then the cloud was really in its earliest stages, I had to go to someone to configure the firewall for me and set up a machine, virtualization, and a lot of the time the storage ran out and you had to fly to London to do it. What year are we talking about? I started there around 2009 or so. And then Ohad did some development, some initiative inside the company, to spin up these environments dynamically and ephemerally. And I was his guinea pig on my team, and we started working together with some product that was, again, pre-cloud, called Skytap, which we used, and it would spin up all sorts of things through virtualization. What does that mean? What does spinning things up through virtualization mean? So back then the dominant thing was VMware, so you'd spin up some instance inside some server farm, and I could get some server and install my application or my data on it for whatever I needed. So that was the first time we dealt with it, and we were even together on this problem. Since then he went somewhere else, opened a startup, and along the way, after that startup, which was on a completely different topic, he started thinking about what he could do and founded a DevOps consulting company. As I said, I kept managing all sorts of teams and I ran into this in every world. And in my last role I was an engineering manager at Proofpoint, and one of the things we had there, we had Terraform and we were on AWS, this is already a bit later in my career. Which is one of the competitors. So Terraform as open source isn't one of the competitors, Terraform Cloud or Enterprise is basically on top of the same framework, so yes, it's one of the competitors. And we had the open source there, the DevOps team worked with the open source, and every time I needed something, or I don't know, to spin up a certain server or change some configuration, I had to open a ticket for them and wait for it to get into their sprint and plan their sprint, and their sprint was once a month, and that was it.

Eyal David: [10:45] So at this point in time, okay, there's a need in the world, your partner at this point has a probably successful consulting business in DevOps, you come along, how did it connect.

Validation with 50 companies and the first pre-seed

Omry Hay: [10:57] And then what Ohad saw was the rise of this world of Infrastructure as Code, and that every company he worked with faced the same problems in all kinds of variations, and the trend of Infrastructure as Code adoption only kept rising. And then he said, wait, I have to do something here, there's some market failure, and he started talking to me, and me with my experience and everything I went through in my career with this whole topic of... You felt the pain. Yes, I felt it in a very, very significant way. He talked to me because we stayed in touch after Ituran, and what happened is that he told me afterwards that he felt he was talking to a user of the system and the platform, because I raised all the pains and all the points and all the problems, both from the technical side and from the product side. More, and that's it, and then we said, wait, let's do something with this, let's go talk to all kinds of people. We started talking to something like 50, 60 companies in the market. How great that you did that. And we got very, very positive feedback, and also along the way a few people said, wait, I want to invest, without us even asking, some small angel investors. And they understood what you were talking about, like at the end of the day it's interesting. Yes, so mainly we went to people who are a bit more high-level, who look at this problem more holistically, or more as how do I improve R&D, how do I help DevOps, all kinds of CTOs and VP R&Ds, Heads of DevOps and so on, and yes, they totally got the point and understood the problem. It's a problem that, you know, exists in almost every company that works in the cloud today. It's also easy to quantify in money. And easy to quantify, say there are places where it's much easier to quantify in money, it's mainly efficiency, meaning how much. Definitely, I'm telling you, but DevOps is a very, very expensive commodity, and we have all kinds of customers who say, yes, you saved us x DevOps people with this product, and all kinds of configurations we do and things we provide that improve their lives and improve efficiency. So it was very, very easy to connect, and then we said, okay, if there's something here that... Wow, what feedback from the market, so you actually got investment offers, soon we'll tell whether you got... Right, it was angel investors who said, come on, I want to put in some money, and then we said, okay, let's try to raise money for this, and we actually raised some pre-seed money from Grove Ventures. At this stage are you working? At this stage I'm kind of half-time, basically what we did is I did all sorts of projects at Ohad's company, and as soon as we got the money we founded the company. It wasn't a very large amount, it was 650 thousand dollars, in today's terms it's a drop in the ocean, and we started running with it and started talking to all kinds of design partners and understanding what the product was. We had a thought. It's funny to think about it, but our first thought was that today there are lots of existing environments, and one of the problems isn't just managing what exists but how do I bring my legacy environment into Infrastructure as Code.

Eyal David: [14:37] Because you identified, sorry for interrupting, but you identified that this is basically the thing you need to start with so that people come in and buy in to the product.

The first bet: converting legacy environments to code

Omry Hay: [14:45] Right, so when we started, adoption really was a very, very big trend, but if you went to a company you'd probably get one of three answers. Either yes, we're all in on Infrastructure as Code because we started yesterday and that's the way to do it, or we have some Infrastructure as Code and we also have things that aren't, and then there's some gap you want. Some kind of maturity model, basically? Yes, right, basically something like that. So you really catalogued them like that. And those who said yes, I really, really want to, but I'm not there yet. And then what we started doing is we said, wait, let's also help them get into this world at all, the market isn't there yet. Let's take existing environments and try to generate Infrastructure as Code from them for the customer. And we started with that, but we also said, wait, but that's a one-time thing that you do, and you also want to manage the day 2 operation, as we call it, meaning I have Infrastructure as Code, the environment exists, okay, what now, now I make a change to this environment, I spin up another service or another application, so the day 2 operation is something we understood we had to do. And I have to admit that very quickly we understood it was very complicated and there wasn't really a market there, and we dropped that story. I have to admit that today we're actually building a feature right now that helps you with exactly that, mapping all the resources that aren't in code and helping you convert them into Infrastructure as Code. Sort of a sitemap of what I have, and sitemap as code. Right, yes, and all with AI and all that fun. So, let me take us back for a second, this is very interesting and also modest, so I'm trying to understand, what year roughly are we talking about for this investment? The investment was in 2019, at the beginning of 2019 we got the pre-seed money from Grove and a few more angels and started running with it.

Eyal David: [16:54] And then you went out with great vibes from the market, great vibes from the investors, everyone got it, you didn't know exactly what you wanted to do, it started to take shape along the way, you started to understand the problems, what is it even like selling to DevOps people?

Who to sell to: VP R&D or Head of DevOps

Omry Hay: [17:07] So I have to admit that at first we also tried to understand who we were selling to and who had the budget, because it wasn't completely clear whether we wanted to go to the VP R&D who says, wait, how do I improve my developers' lives, or go to the Head of DevOps and say how do I improve your life, and not necessarily what we talked about earlier, the self-service, but the day-to-day job of DevOps, which is basically to look after the cloud and understand what's going on, and not necessarily the self-service you give developers. And we had to crack that, it wasn't simple, because at first we went to all kinds of VP R&Ds, and because the product is very, very technical, they don't necessarily know how it works and what the fine detail of that thing is, and when we started going to DevOps people it was much easier. Basically that's the persona we sell to, even if we go for self-service, in the end the ones who implement it are the DevOps people, they have the budget, they understand the problem, they feel it, and it's much, much simpler. And there are all kinds of DevOps people, many of them don't always fully understand the need because they say, wait, I'll build this in ten minutes in my CI or my Jenkins or my whatever. But slowly the market grew and it became a kind of standard. There's also a name in this industry called TACOS that has fairly caught on, and slowly people understand you must have such a thing, because at large scale, when you have lots of infrastructure, lots of cloud, and everyone is moving to the cloud, you basically need something to help you manage the whole thing. So basically at that moment when you understood

Eyal David: [18:56] that DevOps is your persona and that you're going up to that, what did it look like from that moment, like the product suddenly took a certain direction and you understood that's where you're aiming and you're getting traction, how was it?

The pivot to the DevOps persona

Omry Hay: [19:09] So basically in a lot of conversations with customers we understood that we needed to focus on how you solve the problem for DevOps people, how you solve their automations, the tools they're used to working with, say mainly inside my Git, I want everything and I want everything to be connected. And basically there was some, I remember Hunters, who are our customers, there was a conversation with them that Ohad had, and somehow he understood that we needed to focus more on this problem, and that way it would also be easier for us to get into organizations. And we started building a lot of features aimed more at DevOps people and their day-to-day, whether it's automations, policies, role-based access, cost, all kinds of things they deal with, and not necessarily what their developers need. And that was a point in time where we really moved to work much more on those things. So we totally the other side

Eyal David: [20:18] but really all the features and the whole roadmap shifted more in that direction. You followed the oil, and basically what happened from there is that you kept maintaining both of these versions, or did you really focus, how was it?

Omry Hay: [20:34] So broadly it's the same platform, there's simply, in our platform, people who are admins and people who are developers. Even today in the demo we still have a demo of the self-service, and it's something everyone wants to get to, it just takes time until you get there. So we didn't abandon it, and it's the same platform, we just brought much more value in these worlds too, meaning how do I automate all the management of my cloud, how do I know how much this thing costs me before I run it, after I run it, how do I manage it at large scale, meaning I have lots and lots of resources in the cloud that I need to know how to manage, and I want to know who did what and when, I want to set an approval policy. Lots and lots of features simply went more in that direction. We said most of our customers in the end talk to DevOps people, and what matters to them is their day-to-day, what they do. True, it's also important to them to provide services to their developers, in the end that's one of their roles, but we focused on how we get in the door with DevOps people and DevOps teams, and in the end we slowly understand that's the right direction. And from the start was it easy for you to sell the product, or was it something that suddenly, when you went in that direction, you felt something click here, how was it? So I think at the beginning it was harder for us to sell the product, because as I said, all companies today want to move to the world of self-service, and how do I let developers create their own resources, but it's a long and complex process, meaning you have to be ready in terms of your code, in terms of your application, and so it was a harder sales process, and sometimes it's also on the side, meaning it's not necessarily the first pain DevOps people feel. When we made that pivot, suddenly we saw that it was much easier to get in and much easier to start the conversation from a simpler place of, here, I'm solving the things you need. And it was a process that was much easier, much shorter, and basically much smoother.

Eyal David: [22:58] We have a dream, and did it start as a SaaS model from the beginning? Yes, it was a SaaS model, and there was also another pivot there that also opened a gate for us.

JFrog and the self-hosted agent that opened up enterprise

Omry Hay: [23:07] It started with JFrog, who are a customer of ours. We had a SaaS model where everything ran on our side. And then JFrog came and said, wait, that's all well and good, but I can't have everything running at your place. So we built what's called a self-hosted agent, which is basically something that runs inside that customer's cloud account, it can be any cloud account, it can even be on-prem. And it talks to the SaaS, and the SaaS is a kind of control plane, but in the end all the secrets and all the more secured things run at the customer, and it manages that. I remember when we did that, suddenly it opened another very, very big door for us, a segment of companies, yes, another segment of companies that actually need this, they can't even use you if you're SaaS. We also have big customers and enterprises that use the SaaS, but in the end it really opened a door for us, and I remember that after JFrog, even before they closed, suddenly we closed four more companies as customers because of exactly this. The platform and the features were very, very much the same, it's the same thing, it's just where it runs. And that really was another pivot in the company, and today most of our big and enterprise customers use what's called the self-hosted agent.

Eyal David: [24:29] Yes, and amazing, and you also mentioned on-prem along the way, so I understand what other customers you have. Right, yes. Wow, that was really, the enterprise, amazing, and take us to the next leap. So you did it, wow, you got money, I don't know if there was another investment in between, because it was a great period when you could raise a lot of money, but basically you proved things. You say we made pivots, but we proved, we proved, we proved, and we managed to create more and more of a path and revenue. So when was the next leap, like the raise, what was it?

The funding journey: from $650K to $42M

Omry Hay: [25:00] Lots of raises along the way, as I said [unclear], I'm trying to say that you need to subtract 650 from the pre-seed, there's still a long, long way. So basically about a year after the pre-seed we raised a seed, also not very big, of 3.3, after we came out with a beta version open to the public, after we ran the experiments with the design partners. That brought a lot, a lot of interest, both from customers and from investors. Did the design partners become customers along the way, by the way? So some did, partly that's what brought the investment, but that's what brings the investment, yes, basically the feasibility and your ability to turn a design partner into a customer. So some did become customers, and that brought a lot of interest to the company, and the market also grew a lot. In the end there's market education about what this thing is, what's needed and why it's needed. And we did an extension to the seed round, which in total was something like 8 million dollars. Then Microsoft came in as investors in the A round, which was 16 million dollars, and then we did an extension to the A with another fund called Venture Guides from the US, and in total we raised 42 million dollars. In the end, these things are very much tied to your feasibility, how you bring the market

Eyal David: [26:34] to you and how you bring customers and your growth rate, and in the end one of the things is how well you understand the market, that's a significant thing that happened here, probably no doubt. You know, you need to be an expert, especially in this field, especially when you're very, very technical. It sounds like there really were several things that came together, really also the aspect that you understood the market very well, you landed at a great point in time, maybe there's also an element of luck here, but you identified it, and probably the growth in money flowing into tech companies also created more need for DevOps and for convenience, and you caught exactly the needs and kept doing it, simply amazing.

Where env0 is today

Omry Hay: [27:12] And where is the company now? So basically now the company is in a process of relatively accelerated growth, mainly in what's called customers and customer size. We have big customers like Western Union, PayPal, Paramount, and in Israel, companies like I mentioned, JFrog, Redis, Varonis and so on. And basically we understand where the market is going and its growth is very, very large. Soon we'll probably also touch on where this market is and how much it was shaken in the last year. But basically we're there, we're trying to grow significantly, we've hired a lot of salespeople to make this leap forward.

Eyal David: [28:23] Probably worth peeking at your jobs page. Yes. Maybe we'll put a link in the show notes afterwards. So yes, I want us to touch on that, but before, let's just wrap up this part. I just want to know, so what does your day-to-day look like now inside all this chaos you described?

A co-founder's day-to-day and five months of reserve duty

Omry Hay: [28:38] Sure, my day-to-day changes quite significantly. I think as a founder you do what's needed. You still [unclear]. Sometimes, yes, I do panels or whatever is needed for the company to succeed. At the beginning it was mainly around R&D, then product, then customers. Today I'm sometimes in the hat of a sales engineer, sometimes in the hat of customer success, sometimes in the hat of support, whatever needs doing. A lot on the strategic side, meaning what we're doing, who we're selling to, how we sell, how we do things, how we manage the processes inside the company at all levels, in sales, in marketing, in customer success. So sort of the CEO in Israel, like at the end of the day. Right. Again, I think it's more than the CEO in Israel, it's just that as a co-founder you're exposed to everything and need to do everything. That's what I'm saying. Whether it's fundraising or the board and everything that involves. I have to admit that since October 7th I spent five months in reserve duty, so there too life changed from end to end, mainly for my wife but also for the company. And basically since then, so what, were you on the line the whole time, like how was it? So no, I pretty much disconnected. I'm in an infantry battalion, so it's very hard, to the point of impossible, to really manage a routine life with work, and also the intensity was different, not something I know or knew until now, maybe except for Protective Edge. And that also changed my role. Granted, you know, someone had to come and take what I did and step in under the stretcher, as they say. And when I came back I said, wait, there's something here that works, I need to rethink my role and what's more right for the company and what's more right for me to deal with here. And that changed a bit day to day, it's much more go-to-market than R&D or day-to-day management.

Eyal David: [30:57] Maybe that did a bit of good, so it basically says one of two things, probably about you or about the company, either you build amazing infrastructure, or probably both, you have great people there who knew how to step up. I just want to ask for a second, really a question we usually ask, so in a company like yours, what do you think are the most important product skills you need?

The most important product skills in a technical product

Omry Hay: [31:21] Sure, so really our company is very, very technical, and DevOps people, you asked me what it's like selling to DevOps people, so DevOps are people who really love certain things and are very loud, as they say, about voicing their opinion, this is how it should be done and not otherwise. So we get lots and lots of feedback from customers, from prospects, and we really need to distill it. A lot of the time you get some feedback, I need something very, very specific that I want you to add for me, but when you understand what's behind it or what the story behind it is, it's not necessarily the thing you need to build. I think one of the most significant product skills for us is understanding the customer and the story behind what they give as feedback, and understanding how to adapt it so it's more generalist. To generalize. Generalize it, yes, so it isn't specific to that customer, but rather the more correct way to use these frameworks or these worlds or these features, and how to make it accessible too. A lot of the time accessibility is something where you need to understand how to make it accessible correctly. And by the way, we often have two users: one is really the DevOps team or the DevOps person, and on the other side the developer too, how easy it is for them to spin up their environment without needing to know what it is and what goes on behind it. So that's one thing. The second thing, I always say a product person is key in the company, a very pivotal figure, because they need to talk to customers and get feedback, with customer success, with sales, with marketing, for how do I tell the story of this new feature I shipped, and on the other side also with R&D, how you take, okay, here's the feedback, here are the things we want to do, okay, how do I translate it into a PRD or something chewed enough, and how do I bring it down to the ground, because there's always, you know, an existing system or something already built, and how do I make those adjustments and what's right to give up and what's less so. But in the end they're really pivotal, they also touch pricing, and how I price the product so people buy, and how I grow usage to support customer success and the growth of our customers. What's the product strategy, we were just talking about AI, so how do I integrate it into the product. It's really a very pivotal figure in this world, in my view, in any company. I think for us it's even more extreme, because the customers are very, very technical, so you also need to know how to take this technical feedback and maybe use our R&D, and we also have a unique setup in how our R&D works. And basically we need them to be a really key figure in this company. I think those are the basic skill sets, obviously there are other things here to prioritize, UX, all sorts of other things, project management, but in the end that's the essence for me. Yes, yes, I think if I take it for a moment,

Eyal David: [34:47] so I think beyond framing problems and things that come up from customers, which is classic, there's another aspect here, that he needs to know this shit, Omry, right? Because it's a very technical product, on the other side you really have the development team, if you come in [unclear] and understand what you're doing, it won't converge.

When R&D does product

Omry Hay: [35:06] How many product people are you today? So today we have one director of product, until he arrived we did it ourselves, but as I said, our R&D is unique because the product is very, very technical. Our R&D is built, first of all, in a relatively flat structure, and a lot of the things product does, R&D does, meaning a lot of the time a PRD, one of the developers does it, a lot of customer conversations, many people from R&D can accompany or come to such calls to understand the feedback and be more technical and get there. As you said. If on the other side there's a technical person who doesn't respect you enough on the technical side, they'll hear it as who are you and what do you want from me, let me talk to someone who speaks my language. So R&D does a lot in many places, also in marketing, a lot of things, the blog post you publish is very, very technical and you need someone to go over it and see that it really doesn't, that it hits the mark, that it hits the mark, yes, in language, in tone, that we don't do anything silly, don't embarrass ourselves. And basically today the person leading the whole product is our director, but sort of his PMs are R&D today. Interesting, so now let me ask for a moment, just before we move on, what are your challenges today from this aspect? Because I'm sure there are challenges. Yes, so there are lots of challenges in many places, in every company I think there are lots of challenges. As I said, if I look at myself for a second, one is how you make this leap forward at the company level, at the sales level, and bring more customers. If I take a step into the product for a second, then how do you really create this value for the bigger enterprise, and how do you create a product that's more coherent and hits more and more places and use cases, and create a place that's a platform where you manage lots and lots of things. I can give an example, we started with Infrastructure as Code, today we support more things, more frameworks of all kinds, because for example I talked about Terraform, CloudFormation, there's Terragrunt, there's Pulumi, and how do you support that? Because the big ones will have everything, and more than that, today we also support Kubernetes and Helm and Ansible and all kinds of other things, because you want the whole envelope of how you manage all your environments in one place. So this approach isn't a simple thing. Another thing is how you grow and how you do it correctly. As I said, you know, I said I did product until a certain stage, and then it came, you know, the time to bring in someone to do it. Because I don't have all the time in the world to deal with it. And how do you grow the business correctly in general?

Eyal David: [38:17] There's also the aspect here that I think there's this unique structure you described, this culture where R&D also participates in product. The question is, first, how do you preserve it over time, do you preserve it, like, that's probably also totally a thing. Yes, it's one of the things we talk about day to day.

The earthquake: HashiCorp changes Terraform's license

Omry Hay: [38:37] How we improve it, how we improve R&D so they do a better job, whether we need to bring in more PMs to help them with it, but the culture of developers who are very, very, what works has to stay. Yes, a real question how. Okay, great, good answer. Now let's move on to the story, the story you mentioned, we talked about it before the episode, it was amazing, so share it. Sure, so I assume not everyone knows because they don't live in this world, but in August a year ago there was some earthquake in our worlds. Basically, as I said, HashiCorp is a fairly large public company, and soon we'll also touch on the acquisition that happened there. They're basically the father and mother of the most widespread open source in the world, called Terraform, in your world, in our world of infrastructure [unclear], right, and in the world in general. And they decided to change the license of that open source, and it became no longer open source starting from a certain version, and the license basically says that if you have a product competing with one of HashiCorp's products, you can no longer run that open source in your product.

Eyal David: [39:59] Which means that at some point they identified that the industry had grown enough around this thing, and they wanted a slice of the pie, the whole pie.

Omry Hay: [40:06] So they wanted the whole pie, they wanted to sort of kill all their competitors in these worlds. They did it across the board, they have about nine products, all open source, it's a company that champions, or at least championed, the open source model, and they did it to all their products. We look mainly at Terraform because that's our world, but yes, they basically changed it so that anyone competing with their commercial products can't do so anymore. And it was aimed mainly at the Terraform worlds, because that's where there's much more competition and much more financial interest in the end. And another question, was it like ripping off a band-aid, or did you have time to prepare for it? So it was, I'll tell my personal story. It was August, I was on a trip in Italy, and one fine day one of our investors sends us an article that HashiCorp is going to change the license of all its open source. I remember I immediately called Ohad and we started talking and got on calls with the investors, and I didn't sleep at night, and throughout that whole vacation I was mainly in the Dolomites, so on pretty much every cable car I did Zoom calls of all kinds. But yes, it was fast, we couldn't prepare for it, and basically they announced it and said that starting from this version you can't use our products at all, basically from their next version. So it was very, very fast and we had to prepare for it. And what happened in the end is that all the companies that had this problem banded together and founded OpenTofu. OpenTofu is basically a fork, we took the Terraform code and took a right turn. And we started developing on top of Terraform and called this project OpenTofu. Today it's in the Linux Foundation. The Linux Foundation is a nonprofit that supports all the big open source projects, and basically today it isn't part of a company but sort of a nonprofit, which means it will stay open source forever. There are examples like Kubernetes and Istio and all sorts of very, very big things in this world. And today there's very, very big adoption of it, because we proved that lots and lots of the community in the open source world didn't like this move. More than that, your ability to influence the open source today is much greater, there's much more visibility, there are lots of people and community around it, lots of companies that adopted OpenTofu. But at first it was very, very painful and we didn't know exactly how and what to do, and slowly it took shape, and we were in all kinds of conversations with all our competitors about what we're doing and how, and how we preserve this as open source for the community. So that's one thing that happened that really shook our whole world,

Eyal David: [43:21] it also happened in the past, there was Elasticsearch with Amazon and all sorts of other things. I want to, okay, I'm getting the sense here that you're a very modest guy. Okay, so you're in the Dolomites on the cable car, throwing back an espresso, I'm trying, doesn't drink coffee. Doesn't drink coffee, right, we talked about it when I was drinking coffee. And I'm trying to understand, so you got on a call with the investors, the sky is falling at that moment, how did you get from that moment to, if I understand correctly, teaming up with all your competitors and founding this thing, just lay it out for us,

Competitors unite: founding OpenTofu

Omry Hay: [43:55] Of all the arms of this octopus, of these 9 divisions, was this the biggest? They basically have, HashiCorp, two biggest products, one is Vault, secret management, and the other is Terraform. They have all sorts of others, but those are the two big ones, you also see it in the financial reports they publish, and each of them has a commercial offering on top of the open source. And basically what happened, just, I'm only trying to understand, one of the investors threw out the idea, let's all unite. No, it didn't happen like that. We talked to investors, Ohad and I talked, and you know, at first we threw out all kinds of ideas, what do we do, what pivot do we make, or what do we plan next, there were all sorts of crazy ideas. But at some point we said, wait, we have to do something, this isn't the first time this has happened, there's a precedent for this, and Ohad simply picked up the phone to our competitor, to our competitor's CEO, talked to him in the most open way, and then we talked to all sorts of others, we started, we created a Slack channel, at first it was on WhatsApp and then we moved to Slack, and slowly, we first called it OpenTF, then there was some legal issue there. And yes, and it really... Because what, because they tried to go after you guys? So yes, there were a couple, there were two cases where HashiCorp sent each of the companies warning letters of various kinds. One was about, there was a change, be careful not to do this and that. The second thing was really to OpenTofu, there was what's called a cease and desist, some lawyer putting out a letter saying they want to sue us because they think we stole code from their codebase, which isn't allowed. Actually, it was one of our developers who did it, amid all the mess and commotion, and obviously she didn't copy, and we showed how and why. And today we're even more careful than that, we don't necessarily, I'm not allowed to use AI products of various kinds so as not to do that, but yes, there was,

Eyal David: [46:17] We understand they're going to choose. Yes, they tried to protect what's theirs, okay. And how long did it take end to end until you came out the other side?

From petition to fork to the Linux Foundation

Omry Hay: [46:25] So it was around August, very quickly what we did was open a website with a petition to HashiCorp to return it to open source, and if by a certain date they didn't, we'd do the fork. It was a span of something like two months, more or less, two, three months, where pretty much all we dealt with across the whole company was this. And after that we came out with an initial version of OpenTofu, we entered the Linux Foundation, all of this happened, I assume, before October 7th. So less than two months even. So less, yes, something like that, less than two months, like we already had a direction to really get in, and we had conversations with the Linux Foundation and so on, and today it's already, we have... Sorry, and you went to reserve duty, I'm holding off on that, you went to reserve duty for five months. And I went, yes, I was, open source projects of this size have what's called a steering committee, basically a committee that decides on the roadmap, on what gets done, and I was part of it, and because of the reserve duty someone had to replace me.

Eyal David: [47:41] Yes, it was fun and interesting. Crazy, it shows great resilience, in the infrastructure you decided on.

Omry Hay: [47:48] I have to admit that in the end, you know, there are other competitors here who helped us, and we were in the same boat together. That's the infrastructure you built, sounds amazing, kudos, an infrastructure builder, yes, well.

Closing: what Omry wishes himself in ten years

Eyal David: [47:59] Okay, that's a great story. Let's really wrap up the episode. It was very interesting. We've started asking a question here, and I'll ask you too, what do you actually wish Omry in ten years?

Omry Hay: [48:10] Wow. First of all, health. Health, you're already starting to feel it now and then. I don't know, I said I play tennis and now and then I feel the aches. So one, that I keep enjoying, with family, with the work, with life. In the end life is the most important thing, being healthy and happy. That's one of the things, and to keep doing, I think that's what drives you and makes you get up in the morning. Again, it can be in all sorts of ways and forms, not necessarily at work, but mainly to keep doing.

Eyal David: [48:55] How great, we came away with a lot of inspiration. Omry, it was so much fun. Thank you very much. Great, so yalla. Bye for now, and see you in the next episode. Bye bye. Hey everyone, thanks for listening. If you found this podcast valuable, you can subscribe and follow us, of course, for more episodes on Spotify, Apple Podcasts or any other app. Of course, if you didn't find us on some app, I'd be glad if you wrote to us. We'd really appreciate five stars on every platform, and that you follow us, so more listeners can discover us and find the podcast. In every app or on the YouTube channel we have links in the description. So until next time, yalla, be efficient and bye bye.