Subscribe
Share
Share
Embed
The Growth Wizards Podcast is the playbook for B2B GTM leaders who want smarter, repeatable ways to generate demand and build pipeline. Each episode breaks down how CEOs, CROs, and marketing execs drive growth — from top-of-funnel tactics and AI to thought leadership that converts.
Tanner Green: All right. Everyone, welcome to another episode of Growth Wizards. We are very excited to have Josh Epstein here with us today. Josh, thanks for coming on the show.
Josh Epstein: Yeah, I appreciate it. Thanks for having me, Tanner.
Tanner Green: Right on. So Josh, just to start off, why don't you tell the listeners a little bit about yourself, what you're doing, and what you're working on, and the listeners would love to hear.
Josh Epstein: Yeah, absolutely. So I I'm working at a company called Coder. I've been in the software industry about twenty five years now. I think saying that out loud is a little painful. prior to that I was I did a short stint as a technical recruiter. It was my first job out of out of school. And been in been in software ever since, mostly in sales roles, revenue roles. the last year, year and a half at Coder, I've been running a broader part of the business. So I run R and D plus the the go to market teams. So my report to the CEO and I've got a bigger part of the business.
Tanner Green: Love it. Awesome. And when you were how long have you been coming with Coder? Just
Josh Epstein: Little short of three years. So I joined I j
Tanner Green: A little short of three years.
Josh Epstein: I joined at the beginning of twenty four. So early twenty seven will be three years. Yeah.
Tanner Green: Love it. Right on. So can you explain a little bit more about Coder for for the listeners?
Josh Epstein: Yeah, Coder's been an interesting, an interesting ride. So, you know, I joined, like I said, I joined in January of twenty four. We were selling development environments to platform engineering teams, so trying to help companies to onboard developers faster, which at the time the the the hypothesis was tons of value, tons of accelerating the effort of developers. So, you know, it's a space I came from. I was at HashiCorp. For for seven years. That's where I really cut my teeth in leadership. and it was a much harder build than I expected it to be. So Coder early on, you know, was we were trying to figure out what we wanted to be when we grew up. And then middle to the end of 25, sort of the summer of 25, all of a sudden ChatGPT showed up and Claude Code showed up, and all these coding agents showed up in the enterprise. And our customers started to realize that we had a lot more value for them in governing the coding agents than we did in onboarding humans. and so we fairly quickly figured that out. We've got a a really sort of agile team. I'm I'm pretty excited about telling this story sometimes because I realize how quickly we figured out that we were, you know, there's a hockey analogy, right? We were skating to where the puck was gonna be and we didn't realize it. We didn't even know it. So we got really lucky, really smart engineering team here, really creative. And so we've been building this platform we now call the AI operating layer. we've been building that out with partnerships around, you know, very strong brands, a cursor and anthropic and OpenAI and those folks. Every company is using those coding agents, and we have about 200 customers that are running them on Coder. so it's been a really interesting build. you can sort of think about us as as the the the layer that you know governs, secures, protects, and integrates those coding agents, but we're not the actual coding agent. We're sort of that infrastructure underneath.
Tanner Green: Which is what everybody has been looking for and didn't realize they needed it until everybody
Josh Epstein: Yeah.
Tanner Green: Pretty much crapped themselves and you heard all those crazy stories of like
Josh Epstein: Ha ha.
Tanner Green: Explosions on the business losing millions of dollars accidentally, you know? So
Josh Epstein: Yeah, yeah. No, we're you know, I always make the joke infrastructure is the most boring thing until it doesn't work and then you'll pay any amount of money to make it work. So,
Tanner Green: Hmm. Yeah.
Josh Epstein: You know, when Hugging Face got breached, by the OpenAI agent, that was seen as a terrible thing. But if you really look behind the scenes I think it might have been one of the best things to happen to the industry 'cause all of a sudden it woke people up to the risks that that are here and hopefully we can get in front of some of these things. There's a lot of there's a lot of attention in this space right now, that's for sure.
Tanner Green: Totally. Yeah. It reminds me, my wife and I just bought a house a couple months ago and there's lots of stuff going on with it. It's a total pain to, find something and you're like, we just bought this house and now we need to spend, you know, X amount of money to fix this problem that nobody knew was there. Anyways,
Josh Epstein: Yeah.
Tanner Green: But in a house, like there's so many things that happen where you're just like you maybe even know about it, but you're like, it's fine until it breaks, until it explodes, you know, like the electrical we got an electrician come to the electrical box and they're like, Yeah, this
Josh Epstein: Yeah.
Tanner Green: Thing needs to be replaced, otherwise it might be dangerous and it's like, Ugh, that sucks because nothing's broken but
Josh Epstein: Right, right.
Tanner Green: Still works, but like there's a chance that it's gonna totally explode and cause a house fire. So like you gotta do it. So I totally
Josh Epstein: Yeah.
Tanner Green: Get what you're meaning with the with the infrastructure issues. that people feel. It's a it's a the feeling is real for sure. Yeah.
Josh Epstein: Yeah, the only downside to it is is like you said, you only know about it when it stops working or when it's when it starts to hurt. It's the least sexy part of tech there can be. And so one of the guys who runs my sales team for me likes to say, you know, we're you know, we're really focused on our people and our people are companies with serious problems that want to or with serious infrastructure that want to build serious solutions. we're not the it's not the sexy stuff that everybody wants to talk about, which is the developer experience and that that you know that that UI that the coding agents offer. So we're we're there when you need us. We're the but it's it's not sexy for for most people.
Tanner Green: Yeah, but definitely needed. So so I love that. so let's jump into a little bit of news. I'm curious to get your thoughts and then we can go jump back into the the coder side of things. But
Josh Epstein: Yeah.
Tanner Green: I saw an article recently that Meta, formerly Facebook, launched a business agent that opens its ads platform to AI agents through the MCP. And it's not necessarily this specific instance of what's happening in the world that everyone's kind of releasing MCPs and it's all getting attached to your Claude instance or to your ChatGPT, whatever it is. And
Josh Epstein: Yeah.
Tanner Green: And it leaves a lot of capabilities for us to have
Josh Epstein: Really?
Tanner Green: These AI agents kind of meddling their hands in every piece of our technology. And so being an infrastructure guy
Josh Epstein: Yeah, yeah. Good thing.
Tanner Green: With Coder. what are some of the positives here? What are some of the risks that all these AI MCPs are communicating with each other and everyone's opening up these doors to AI agents? I gotta imagine that it's gonna be kind of a a crazy spaghetti bowl of network issues that are going to blow up in somebody's face. It it's gonna be cool for the go-to-market side, which it's cool to see that you're in the go-to-market side for an infrastructure. So this is a perfect realm for you of like the GTM side, it can there's a lot of upside to it, a lot of scaling capabilities and a lot of potential there. But what do you see in this space?
Josh Epstein: So I actually see what you said a little differently than you said it. so fundamentally, I see this as sort of an unraveling of some societal norms that have protected us for as long as I've been alive. And there's this sort of natural unease with other people being able to see certain pieces of our information, right? So, you know, you're not posting, your journal online, you're not posting your your bank account online, you're not posting your email online. And I was raised with this moniker that, don't write anything that you don't want to see on the cover of the New York Times. And in this moment in time, the d what you're describing People are basically putting their email out on the internet and putting their, you know, their bank account number on the internet and putting all of the they're tempting these agents to go explore and find because all it is to an agent is data. All it is to an agent is I was prompted to look for this thing. I'm gonna use any way I can to look for this thing, whatever it is. And so if it has access to your email, it's gonna go into your email. If it needs to expose something from your email, it's gonna expose it. If it has your bank account information or it gets it through your email, or it gets it through having access to your phone, I don't wanna quote the guy from Terminator, but it will not stop until it gets what it's trying to get. and so I just think the whole idea of of caring about our privacy and caring about our data at an individual level is really important. and I I worry that it's that's getting corrupted because it's so easy to give Granola access to your email so that it can write an email to somebody after you're done with the meeting. But now you've given AI and its LLMs access to data that it didn't have before that it's going to use to train itself on. And there's no reason why if it finds something that it can use, it's not going to use it. in a way you didn't intend it. And so my bigger response to what you said is I think there's gonna be some really bad things that happen to people, normal people that are don't realize what they're doing with their information before we start to peel this back and put better protections in place and, you know, literally get to the point where there's legislation to protect people from this kind of thing. But I think some really bad things are going to happen before that happens because we're so nonchalant about sharing This information. And as an older guy, I'm on the other side of the generational divide to this. I don't give AI access to anything of mine, anything personal. and I will happily handwrite every email and handwrite every thing that I need to. I use Granola to take notes and then I read the notes, and then I use my brain to do the thing that I need to do with the notes. It's just I'm trying to put a human swivel chair in between. one system and another so that my data and my bank account information and my social security number and all that stuff doesn't get accidentally leaked out into the world. Because I I just I think that's coming, that reckoning is coming. So
Tanner Green: Mm-hmm. Yeah. So I totally agree. I think there's some even bigger issues that are coming to fruition soon, you know, and they're already happening, honestly. And in your side at Coder, tell me the story that you guys portray in the space of warning people, because Coder is Like you said, infrastructure, it's hard to like warn people and tell that something's gonna be wrong until it actually does,
Josh Epstein: Yeah.
Tanner Green: And then they're like, crap, well now we're screwed. So so what is
Josh Epstein: Yeah.
Tanner Green: Your guys' story? Like, how do you portray that in the go to market space?
Josh Epstein: So it's an interesting aside because the question you asked before, there's a lot of companies that are foregoing traditional security practices and and governance practices and letting the efficiency and effectiveness of AI tools sort of trump that, right? There's a decision to be made. Do we want to let our everybody use these tools to be that much more effective? And I you know, don't get me wrong, I think AI is extremely effective at certain things, right? it writes bad code, but it writes code. It you know, it writes decent, English prose, but I don't think it thinks very well. I don't think it replaces thinking. And so, people who use it to to make decisions for themselves, I think, are are doing themselves a huge disservice. when we get in front of customers, the ones who have traditional security teams and compliance teams and things like that, they are naturally pushing back, right? They are saying, hey, we're we love this, these tools, we want to give these tools access to our customers, but we need to figure out a way to protect. and so we're getting a lot of love from the security. The CISO is becoming a a friend of Coder. We have you know multiple, and by the way, this has never happened to me before in 25 years. I ask for time with the CISO and more times than not, I get an enthusiastic yes, and we have conversations about it. So the feedback from them is that there have been multiple issues in in their companies that they haven't publicized of agents doing going rogue and doing bad things and so they're looking for ways to protect themselves. I know a Fortune ten bank that has five figures of one of the coding agents. I'm not gonna out the bank and I'm not gonna out the provider running I'd like at some point to do business with them so I wanna can protect their their identities but
Tanner Green: That's a good idea.
Josh Epstein: Running almost 10,000 instances of one of these coding agents on laptops with no governance. And they're in a lot of trouble because their auditors are aware of this and are and are moving to end what was otherwise a massive strategic program for this bank. And you know, 10,000 developers is not it's not even 10% of their population. But the risk that they put themselves and their customers on, and I have a mortgage with this bank. I almost tried to cancel it for fear of my information being exposed. that's how personal this is to me is knowing these banks are doing this and they're not doing it on purpose, but they didn't realize the risk that they were putting their customers under. so what I'm getting at is we almost don't have to storytell. They've done they've got the the story down. They understand what a breach at this magnitude would mean to them and their customers. the federal agencies are doing a great job of pushing back on the coding agent companies, especially the ones who are cloud hosted. and they're just saying, look, we're not gonna, FedRamp is the lowest level of security or of certification we would allow. Even then, in a lot of cases, they won't let these things run in certain parts of the environment. And so more and more having a l legitimate governance and infrastructure control story like Coder, is you know, we're not the only player in the in the space. So we're you we're excited about the market that that's showing up and and the way the customers are responding. And you know, these these deals we're doing are the you know, we when I got to Coder, our average deal size was in the low five figures. it's now in the mid six figures and growing every every quarter. So
Tanner Green: Yeah.
Josh Epstein: It's been really and and customers are doing three year deals and you know, they're they're buying for growth because they know where they're gonna go. this is an inevitability in most companies, most large enterprise customers. And it's really nice to be sort of in the right place at the right time with the right solution, even though it's took us two plus years to get here, two plus years of
Tanner Green: Yeah.
Josh Epstein: Of of pushing the rock up the hill to get here. So you know, I'm looking forward to the next year, two years of growth that that we're gonna have.
Tanner Green: Yeah, man, that's awesome. And I think it's cool that you recognize like, hey, you're piggybacking off of the people who are already aware of this issue and who agree with you, you know, who are the champions inside and also grateful enough, the decision makers that you're trying to that you're trying to connect with are those are the ones that are concerned about this already. And they're like, and they've seen the horror story. So really it's like the misfortune of others is kind of the benefit to Coder. at some point, which is which is not necessarily like taking advantage or anything like that. But it's just like like you said, it's the right place, right time. and like, you know
Josh Epstein: Yeah, yeah. We're and we're not we're we're not piggy hopefully we're we're not piggybacking off of the bad fortune of others. You know, we Coder's
Tanner Green: Yes.
Josh Epstein: Always been an open you know, we've always been an open source company from the beginning. So companies that aren't that don't need all of the bells and whistles can s can use Coder in their in
Tanner Green: Yeah.
Josh Epstein: Their environments and you know we we definitely are going down that path. But yeah, this is this is a real problem. We're excited about the ability to help companies and partner with you know, the other thing which I've just because I'm on this topic We're partnering with the leading coding agents. So Cursor and and Anthropic and and some of these other folks, they're leaning in with us on making sure that their coding agents work well in the in the Coder environment so that they can go to large companies and say, Hey, we're gonna partner with so the infrastructure becomes coder infrastructure running on their environment and the coding agent and the LLMs are provided by those players. So right around the time that this thing launches, we're gonna be launching a couple partnerships with Coder, with some of these coding agents that we talked about.
Tanner Green: Yeah, I love it. Yeah. And to clarify, piggybacking off the misfortune of others is not what I meant at all. It definitely was meaning
Josh Epstein: Ha ha ha.
Tanner Green: Meaning the joining like minded people in the concern for the welfare of humanity from AI. How about that? That sounds that sounds a little bit better. A little bit more diplomatic,
Josh Epstein: Well done. Well done. Well done.
Tanner Green: If we'll say that that that way. well
Josh Epstein: Yeah, no, we're yeah, I appreciate it.
Tanner Green: Yeah, so definitely definitely on board with that. I love what you guys are doing. Honestly, it's so cool to see that. So I love the the partnership side with these people. It's almost as if, you know, I think one of the biggest things that I am a fan of is in the California gold rush, you know, one of the the most successful people always built the tool for the gold mining and to hand off
Josh Epstein: Yeah.
Tanner Green: To the rest of everyone and let them, you know, try and find the gold. And great if you did, but if you didn't, you know, you're out of luck, right? And so
Josh Epstein: Yeah,
Tanner Green: When you join the
Josh Epstein: Yeah.
Tanner Green: Tool side, like you're joining in the partnerships with these LLMs. I mean, that that's just like so much security and so much stability for a company involved with AI as a whole. And I think that's a really smart
Josh Epstein: Yeah, yeah.
Tanner Green: Move and really cool. So, how did you guys go about
Josh Epstein: Yeah.
Tanner Green: Doing some of those partnerships? Was it I mean, you said right place, right time, but there's gotta be more than
Josh Epstein: I yeah, there's always I mean we're always thinking about how we can how we can get in front of our customer you know, more customers faster and you know, it it's all about awareness and and and act and and sort of access. and I've been doing this for a long time. by the way, what you said, I just want to call it out. my CRO c you know, all the time. We're in the picks and shovels business. That's what we do, right? We sell picks and shovels. So he's right in the same thought process that that you're in. Partnerships are interesting, right? So there's two that I'll talk about because they're both I think we're we're in good shape on both of them. Partnerships are weird because you have to time them up and if you announce something early, it can hurt something. But we've been working closely with the team at Cursor that got acquired by SpaceX and they're really focusing on improving the the posture, the security and governance posture for their customers. And so That conversation has been going on for a couple months. it started out because we have friends in common who just said, Hey, I think you guys are are thinking about this the same way. And so I went and sat down with their leadership team. they have an unbelievable, almost uncapped opportunity because the the you the user experience for their coding agent is that good, right? And so they're just trying to figure out how to remove as much friction from that part of the business as they can. And we are that way to for them to remove friction, right? We're not the only way. They can there's other players in the space, but we're definitely focused on the largest enterprises and federal agencies that they want to play in. Anthropic is a little different. Anthropic has been a customer of Coder's for two plus years. Every developer at Anthropic uses us, they've been public about it. and so they've been just a phenomenal customer and they've helped us to really understand what the market needed in this space and and And then we started to get customers that would ask us how to deploy Claude Code in the enterprise. And so that's where Coder really started to get the lift. That early lift that I talked about was in that space. so they came about in different ways. Cursor and a more traditional, hey, I see a a space for both of us. Do we want to partner? And Anthropic was using us in in their environment, large quantity of their developers, our Coder users. And then one of our investors was very quick to tell us that they basically made Coder a requirement for every single developer in their organization and they pushed us down this path. So they wanted us to really reinforce the technical integration with Cursor and Anthropic. so it you know it was sort of all three of those things coming together at the same time. it also didn't help Tanner that we were having a run as a company. So people know who we are now. We're not, you know, we're not a little tiny company in the corner. We've, you know, our business is built up, our market awareness is built up. so I think a lot of those things all came together at the same time. And we're having we're having that sort of growth moment that combines all these. I don't know whether Cursor and Anthropic would have partnered with us two years ago the way they are today, just because of our the size difference. Yeah.
Tanner Green: Yeah, that's fair. Well, well, I love what you guys are working on. I love how you guys are going about it. It's it's a ton of fun. I I wanted we're coming up on the end here. I wanted to ask you one last question. What is your team doing specifically with AI in the go to market strategy or already working?
Josh Epstein: Yeah, so that's a whole nother podcast.
Tanner Green: Ha ha ha.
Josh Epstein: I have a a phenomenal customer solutions team. And one of the guys who runs that for me, his name is David. He's been building the go-to-market brain for us for the last 18 months. And it, you know, we've gone through a couple He's a big fan of the AI DLC, the AI development lifecycle. So he he you know he uses that for his his teams that are going to market with customers. So these are the SEs and SAs and FDEs and field CTOs that are working with our customers from a technical perspective. He's taken the AI DLC and applied it to their work and how we learn from our customers. So when a customer talks to a product manager or a sales team, Or it's going through a POC, all of that information gets captured with the customer's permission and analyzed through an AI, a series of AI. And we use it to understand the opportunity to put features into our into Linear, which is our tool for managing feature requests and giving support engineers an opportunity to create knowledge base articles when questions come up about issues. all the way through to analyzing is this a real opportunity for Coder? Do we want to invest time in it? what are the, you know, it does a MEDDPICC analysis so we know where our gaps are. It it does I'm a big org chart and opportunity review nerd. So it'll create an op an account plan, an opportunity plan for the seller so that they can face off with their leadership and really review where the opportunity is. So he's built a very robust system. using AI and you know using our internal turn internal data so much so that he ended up hiring somebody full time to manage it for him so it's become an internal app that does an enormous amount of work for us. But when I need to ask a question, instead of trying to figure out who's the seller on an account, who's the SCE, who should I be asking this question to, I just go into the tool, ask it the question, it goes and searches the system, all the data that we have and gives me an answer. gives the sellers what they need to run a sales campaign, probably I don't think we've quantified it, but my guess is it saves you know, every seller in the company and every executive in the company five to ten hours a week of time because it does all this work for us. So it's an an absolute insane ROI for us. But it also took six months and a lot of, sweat equity to get there and and David's put a lot of work into it. had to retune it at one point because it was going down a a different path and we realized there was a lot more value in the path we went down. So we did the AI DLC prototype and fail, prototype and fail a couple times before we got it right. So
Tanner Green: Yeah. Which I think honestly every company needs to go through at least once to figure out how to put AI. Yeah.
Josh Epstein: We have to know what our customers are going through. If if we don't have empathy for what they're going through, then I don't think we can be a good partner. Yeah.
Tanner Green: Yeah, a hundred percent. I love it. Well, awesome. Josh, this has been a ton of fun, man. This has been really interesting, fascinating conversation, really relevant to today. I love it. how can people go find you and connect with you?
Josh Epstein: Well, I mean I'd say f first thing is if if I hit home, go to Coder. I'd rather you win go to coder.com. I'm boring. I'm just you know I'm just a sales guy, but I'm on LinkedIn if you wanna reach out. but I'd be most interested in, you know, what people think about what we're doing and and where we need to take the solution as we as we move forward. 'Cause that's the other thing about this AI world. it's changing very quickly. So We're already thinking about what we need to be building for twenty seven. So
Tanner Green: Yeah, I love it. Awesome. Well, Josh,
Josh Epstein: Awesome.
Tanner Green: S thank you so much for coming on the show and thanks for the listeners for tuning in.
Josh Epstein: Yeah, appreciate it. thanks everybody.