How Guild AI discovered their real buyers weren't developers — and rebuilt their GTM motion around CIOs and CSOs | James Everingham

Summary

---

BUILDERS Podcast Episode Summary

Episode: How Guild AI Discovered Their Real Buyers Weren’t Developers—and Rebuilt Their GTM Motion Around CIOs and CSOs
Host: Front Lines Media
Guest: James Everingham Steven (CEO, Guild AI)
Release Date: July 17, 2026

---

Episode Overview

This episode of BUILDERS features James Everingham Steven, a seasoned five-time founder and current CEO of Guild AI. The conversation dives into James’s journey from his early days building technology startups in the 1980s to the recent pivot at Guild AI—transitioning its go-to-market (GTM) approach from targeting developers to focusing on organizational leaders like CIOs and CSOs. James discusses learning from customer needs, navigating rapid industry changes, and building thriving teams and cultures in tech.

---

Key Discussion Points & Insights

1. Early Entrepreneurship Roots and Lessons (01:17–04:00)

• Challenging Beginnings: James shares anecdotes from founding his first company in central Pennsylvania during the late 1980s, struggling to access capital and break into tech while far from Silicon Valley.
  • “I think the first time I walked into a bank, I had shorts on and asked for a $25,000 loan and they laughed at me. But I think 13 trips to the bank and I finally got my act down.” (01:29)
• Initial Tech Exposure: Tech wasn’t a typical career path—more “blue collar” than prestigious or impactful then.
  • “My mother would be very upset thinking I was spending all my time on this video game, she called it.” (02:28)
• Motivation: He followed passion and real-world problems rather than abstract business theses.
  • ”I did try to start a company once because I just wanted to start a company and that was my failure.” (03:25)

2. The Genesis and Evolution of Guild AI (04:00–07:58)

• Identifying the Problem: Experience at Meta, supporting over a thousand engineers, unveiled the complexities of managing AI agents at scale.
  • “Once agents start doing real work inside your infrastructure, they need to be managed. They need a way to control access, to have auditability, to have security. They’re a lot like humans in a way that they’re non-deterministic.” (05:08)
• Agent-Centric Security: Describes the shift from defending against external attackers to managing risks from internally-deployed AI agents.
  • "If you think about why VPNs and the cybersecurity software business popped up… now we have a different problem. When you start using AI, you have these unintentionally bad actors possibly trying to take down your infrastructure from within.” (00:00/05:36)
• Product's Breadth: Although initially focused on developer productivity, Guild AI found broader applications (legal, marketing, HR, finance) resonated more with design partners.
  • “That was a, an interesting observation for us, having focused on developer productivity for so long…there’s probably more draw for agentic workflow optimization in those areas and even the software development right now.” (07:37)

3. GTM Pivot: From Developers to CIOs/CSOs (08:25–09:40)

• Initial Go-to-Market Mistake: The team’s inherent developer bias led them to focus on a developer bottom-up approach, but enterprise leaders were actually driving the decisions.
  • “We were trying to build a product that would appeal to developers. I believe me…But it turns out that… CIOs, CSOs and CTOs are really the ones who are worrying about this more.” (08:25)
• Current Sales Motion: Direct, top-down conversations are required due to unfamiliarity with agent management and AI productivity within large organizations.
  • “I don’t like tops down, but I think it’s kind of necessary in today’s market when people are trying new things they’re not familiar with and it’s on an accelerated timeline.” (09:16)

4. Selling to Engineering Leaders: Pitfalls and Best Practices (09:40–12:54)

• AI-Generated Outreach: James notes it’s easy to spot and ignore mass AI-generated outreach; genuine, informed engagement is essential.
  • “Spotting outreach that was written with AI is pretty easy… Not really understanding the value proposition or what our business does is…drives me nuts.” (10:01)
• Do Your Homework: Understanding platform realities and unique company challenges trumps generic pitches.
  • “Understanding like deeply like what the barriers and the friction is in the company that might either make your product work or not work is pretty key.” (11:32)
• Collaborative Approach: Openness to learning and partnership is preferable to overconfident “we can solve everything” pitches.
  • “People coming in and saying like, we can solve your problem, I'm always skeptical of, but people saying, look, I don't know, but like I want to partner with you…I love the ability to try to solve this together.” (12:28)

5. Product-Market Fit and Customer Signals (12:54–16:28)

• Current Stage: Guild AI is newly launched; confident but not yet claiming product-market fit.
  • “Right now we're just going to market. So like, I would say we don't [have PMF]…we’re literally weeks out of the gate from public launch.” (12:57)
• Strong Signals: When products go viral, users clamor for them, or complain when withdrawn, that's PMF.
  • “A non-obvious one…When you take it away, people complain...We learned what really worked because people would complain if it went away.” (14:18)
• Iterative Product Philosophy: Encourage early user exposure and adaptation rather than perfecting in isolation.
  • “Design to 50% debug into existence. Relentlessly measure how your users are using the software. Don’t fall in love with your idea.” (15:46)

6. Company Building: Team, Culture, and Vision (16:28–19:58)

• Team Building: Majority of James’s efforts focus on hiring stellar talent and aligning teams directly with customer problems.
  • “A lot of my time right now is team building and getting the right team in place. So hiring, getting the best people in the industry.” (16:37)
• Culture: Emphasizes kindness, collaboration, diversity, and customer obsession; prefers outcome-focused, bottom-up teams.
  • “As a leader, you should be very public about what your values are. I like building teams where we're all kind and collaborative and value the same things.” (17:41)
  • “Building great software is an act of empathy. So like, if you feel the pain yourself, you’re going to come up with a better solution.” (18:38)
• Vision for Guild AI: Predicts a world where companies run thousands of agent processes, requiring a dedicated management layer.
  • “We firmly believe…agents are going to be prevalent in the infrastructure…hundreds or thousands...So if you can envision that world, then it's pretty obvious you need a new type of agent operating system or infrastructure layer…” (19:06)

---

Notable Quotes & Memorable Moments

• “Design to 50% debug into existence. Relentlessly measure how your users are using the software. Don’t fall in love with your idea.”
  — James Everingham Steven (15:46)

• “If you think about why VPNs and the cybersecurity software business popped up… now we have a different problem. When you start using AI, you have these unintentionally bad actors possibly trying to take down your infrastructure from within.”
  — James Everingham Steven (00:00)

• “I’m not a big mandate AI usage person, so I don’t like tops down, but I think it’s kind of necessary in today’s market when people are trying new things they’re not familiar with…”
  — James Everingham Steven (09:16)

• “Building great software is an act of empathy. So like, if you feel the pain yourself, you’re going to come up with a better solution.”
  — James Everingham Steven (18:38)

---

Timestamps for Key Segments

• [01:17] – Building startups in the 1980s and following passion
• [04:00] – Transitioning to founding Guild AI from Meta
• [05:08] – Defining the agent management/security problem
• [07:37] – Realization: Real buyers aren’t developers
• [08:25] – GTM motion pivots to CIOs/CSOs
• [10:01] – Bad sales habits: spotting AI slop outreach
• [12:57] – Judging product-market fit
• [15:46] – Building products: release early and measure
• [16:37] – Focus on hiring, team building, and culture
• [19:06] – Guild AI’s vision and future

---

Final Takeaway

James’s journey with Guild AI underscores the importance of truly understanding one’s buyers—even if it means pivoting a GTM strategy anchored in personal identity as a developer. Built on lessons from experience at Meta and multiple startups, Guild AI is tackling a new class of security and management challenges for enterprises adopting AI agents. The episode is rich in practical insights for founders and GTM leaders alike—especially those navigating rapidly evolving markets and customer needs.

---

Careers and Product Info:
Guild AI is in open beta and actively hiring; details at their website.
[19:58]

---

Transcript

[James Everingham Steven]

If you think about why VPNs and the cybersecurity software business popped up, there was intentionally bad actors out there trying to compromise your company infrastructure. It's like now we have a different problem. When you start using AI, you have these unintentionally bad actors possibly trying to take down your infrastructure from within.

[Frontlines IO Host]

Welcome back to another episode of Builders. As always, this show is brought to you by Frontlines IO, Silicon Valley's leading B2B podcast production studio. If you're bringing technology to market and want to learn from your peers, we have a library of more than 1200 interviews with Venture backed founders and marketers. Where they talk, all things go to market. Of course, if you want to launch your own podcast, we offer podcasts as a service to more than 80 tech startups. The idea there is very simple. You show up and host and we do everything else. Now with all that said, let's jump into today's episode.

[Podcast Host]

Today we're speaking with James Everingham Steven, CEO of Guild AI. James, welcome to the show.

[James Everingham Steven]

Thanks for having me on. Excited to be here.

[Podcast Host]

Of course. Looking forward to this conversation. You're a five time founder, you have a lot of history, a lot of things that we can talk about. I want to talk about something very specific and I want to go back to 1987, the founding of your first company. What was it like to be building technology companies in the late 80s?

[James Everingham Steven]

Well, you know, it was actually pretty challenging. I was pretty young, I didn't really know what I was doing. And to make it even harder, I was in central Pennsylvania, was in Silicon Valley. So finding capital was a struggle. I think the first time I walked into a bank, I had shorts on and asked for a $25,000 loan and they laughed at me. But I think 13 trips to the bank and I finally got my act down. It was harder. I mean, technology, the market was smaller, it was more complicated to build things and to get things going. You know, my first company, the reason I even started it is because I had some open source software. Back then it was called freeware or shareware. And you'd basically say, hey, if you find value, maybe donate or send me some money. And people were sending me a lot of money. So I'm like, well, hey, you know, maybe I should try to turn this into a company. So it wasn't a thesis, it was also a working product that I had to tee off of to get that going.

[Podcast Host]

I feel like now tech is very mainstream. It wouldn't surprise me, someone from Pennsylvania or Wisconsin or like Zimbabwe is into tech now, but back in the late 80s, like you said, being there, what was your exposure to tech in the first place?

[James Everingham Steven]

So, you know, I didn't even think tech could be a possible career path when I was growing up. You know, it was computers. I didn't really know much about them. I started using them early in high school. My mother would be very upset thinking I was spending all my time on this video game, she called it. So it just wasn't a career path. It was actually even more like a blue collar work back then. You know, my first job, which was before I started my company, was actually working at the university at Penn State. And you know, I had to be there at 8am I had to wear khakis and a dress shirt. I had to stay till exactly 5pm it was like a very different, it was a very different career back then. It had much less impact. Once again, you know, there wasn't a lot of computers, so it was just completely different.

[Podcast Host]

And when you take a look at these five different companies, is there a through line? Is there something similar? Is there a pattern?

[James Everingham Steven]

I would say the pattern for me personally there has been following what I'm passionate about. And you know, I never really set out in my mind with, look, I want to start a company. That wasn't always my goal. My goal was how do I. I've had an observation that I think is important, that I could build something interesting and I want to get it out to the rest of the world. I did try to start a company once because I just wanted to start a company and that was my failure.

[Podcast Host]

I think that says something. Let's talk a little bit about what you're building today. And I want to go deep on the product, but maybe let's talk about how you decided to make this the fifth company. I'm sure if you're like a lot of other founders I know, myself included, I have a long list on the notepad of ideas, things I could go after. What was it about this problem that made you say, yep, that's it, number five, let's go?

[James Everingham Steven]

Yeah, I think, you know, one, it was a culmination of things that came together all at once. I think, you know, when I was spending my last, the job that I had left as I was supporting the developer infrastructure teams at Meta, you know, it was a pretty large team, over a thousand engineers. We build all the developer tools. We built something that was pretty exciting internally where we learned a lot of lessons about like what happens when you start to scale agents inside a large company using that lens and having like the entrepreneurial bug in me. When you see something working well for something that sort of mirrors product market fit for the world. You know, if you think of Meta as a cloud company with a bunch of businesses built on top of it, WhatsApp, Instagram, Facebook, you know, Oculus, you see that, you know, it gives a lot of conviction. So I didn't think this existed yet in the world. The solution? I had a lot of confidence that a lot of companies were going to run into this problem, which I think we're seeing now. And the market timing was great. And I wanted to build something for 45 million developers, not 45,000. So that's the reason I went and started GIL.

[Podcast Host]

And how do you articulate the problem?

[James Everingham Steven]

What we build is infrastructure for managing and deploying agents in a company. And once agents start doing real work inside your infrastructure, they need to be managed. They need a way to control access, to have auditability, to have security. They're a lot like humans in a way that they're non deterministic, right? You can't predict what they're going to do. So if you're going to have them doing things directly in your infrastructure, you need controls around them. Another simple way that I like to describe this to people is that if you think about like why VPNs and the cybersecurity software business popped up, there was intentionally bad actors out there trying to compromise your company infrastructure. It's like now we have a different problem. When you start using AI, you have these unintentionally bad actors possibly trying to take down your infrastructure from within. So you need a layer to help manage and protect from that.

[Podcast Host]

When you think about the competitive landscape, we don't have to talk about competitors specifically. That's kind of irrelevant. But like, how do you think about the buckets or the different categories that you compete with?

[James Everingham Steven]

I think, look, there's a few different categories. There's people that are trying to build agents and there's like agent building platforms. There's frontier model companies that are also building a lot of these tools, but they're very specific to their stack, so they're not neutral. So you get sort of tied in to a frontier model solution if you go with that. And it's also sort of not their main focus, right? So like our main focus is governance, auditability, security, cost controls, all of that stuff. So I would say that like we believe we're the first company that's building this very specific Core competence. And it's going to be a really needed core competence. That's where we have experience is building this stuff at extreme scale in regulating companies that need compliance. So that's what we're bringing to the table here.

[Podcast Host]

And is it developer productivity? Is that the category? Is that the line item or what's the line item that customers are buying?

[James Everingham Steven]

Yeah, I think that for customers right now, you know, we started this developer productivity mainly because we're developers and we like to build stuff for ourselves, right? And so like, we started this company going like, all right, what do we want to build for ourselves? As we started exploring with our first design partners, it turned out that, you know, we realized what we built is basically a pretty generic solution that like works for pretty much any workflows in a company, not restricted to software development. In fact, most of our design partners are looking at things outside of that. And that was a, an interesting observation for us, having focused on developer productivity for so long. So marketing, legal, hr, finance, there's probably more draw for agentic workflow optimization in those areas and even the software development right now.

[Frontlines IO Host]

This show is brought to you by

[Podcast Host]

Frontlines Media podcast production studio that helps B2B founders launch, manage and grow their own podcast. Now, if you're a founder, you may be thinking, I don't have time to host a podcast. I've got a company to build. Well, that's exactly what we built our service to do. You show up and host and we handle literally everything else. To set up a call to discuss launching your own podcast, visit Frontlines I.O. podcast. Now back to today's episode and is it top down, bottoms up? What does the go to market motion look like?

[James Everingham Steven]

So we're early in ours. You know what we started with both like, you know, we were trying to build a product that would appeal to developers. I believe me. And you know, using the lens of myself being a developer, I feel if I build something useful for me, it might be useful for others. I think all developers sort of think that way. But it turns out that like, the problems that we're solving, so CIOs, CSOs and CTOs are really the ones who are worrying about this more. So right now, you know, we're having direct conversations with like those types of leaders in the company and it starts out to be somewhat tops down. And the reason being is that people don't really know where to start with AI productivity and they don't have a clear starting point. So what we learned is that if you put the right Tooling in and you put the right services in and you start tops down and you can demonstrate value, then those tools earn the usage from the developers over time. So I'm not a big mandate AI usage person, so I don't like tops down, but I think it's kind of necessary in today's market when people are trying new things they're not familiar with and it's on an accelerated timeline, you sort of have to start tops down.

[Podcast Host]

When you were in your role, VP of Engineering, you know, thousand developers, I'm sure that there were a lot of vendors trying to sell to you on a regular basis. What would you say were like the top things that would just drive you crazy or the things that other startups that are listening and are founders that are listening in should absolutely not do when they're trying to sell to a VP of engineering at Meta or these types of organizations?

[James Everingham Steven]

Yeah, I mean, I would say that, like spotting outreach that was written with AI is pretty easy. So like, if I get an email paraphrasing the marketing language on my site, repeating it back to me like that just drives me nuts. Not really understanding the value proposition or what our business does is somebody showing a deep insight to what we're doing gets a lot more attention. So I would say those are a couple of things that sort of drive me nuts. Yeah, I think that there's a lot of people out there right now using AI to sort of do shotgun outreach and, you know, and they're attempting to make it look personalized with AI. And I think that, like, that becomes pretty obvious pretty quick that that's what's happening. It's reminding me of the first banner ads on the Internet. And like, those were very effective for about three months and then they weren't. I think that's what's going to happen with this AI outreach right now.

[Podcast Host]

Yeah, I mean, it's, it's brutal right now and I think it's always very obvious that it's AI. You can tell. I don't know what it is, but you can always just tell. It's like just very, very clear. It's an AI slop email.

[James Everingham Steven]

That's right. And it's coming in like, I probably get a hundred of them a day right now. So yeah, it's hard to sort through.

[Podcast Host]

What about in the previous role though, when you were at Meta, you know, running this big engineering team? I'm sure that there are vendors there that were coming after you or you're trying to sell it to you. Like, what would they get wrong. And the reason I asked you, a lot of founders are listening in and you're probably trying to sell to the metas and market to a VP of engineering. So maybe just speaking to them directly of like, here's how you could have marketed and sold to me better back in my previous roles.

[James Everingham Steven]

Yeah, for sure. I think that the value proposition that people will pitch to you doesn't really potentially map up with kind of even feasible on like the platform that you're working on. So understanding like deeply like what the barriers and the friction is in the company that might either make your product work or not work is pretty key. I can give you an example, like cursor, like cursor came in and was like you should all use cursor. And ultimately by the way, we did end up using cursor in certain areas. But look, you know, billions of line monorepo like bespoke languages that don't work with frontier models. You need fine tuned internal models like, you know, it just doesn't work. What something like 16,000 internal tools at Meta, like it's massive. But coming and just saying like you should all use cursor and not understanding like what it would actually take to get that in is probably a mistake. Companies that come in and they do their homework and they're like, hey, here's some things we'd like to partner with you. I also think people coming in and saying like, we can solve your problem, I'm always skeptical of, but people saying, look, I don't know, but like I want to partner with you and like I love the ability to try to solve this together. So people, you know, sort of admitting where they might fall short or what they don't know is also like refreshing.

[Podcast Host]

Do you feel that you have product market fit now?

[James Everingham Steven]

Right now we're just going to market. So like, I would say we don't Again though, we're using lessons that we learned that had a product market fit. So we have a lot of confidence that we will. But you know, we're literally weeks out of the gate from public launch.

[Podcast Host]

When you think about knowing when you're a product market fit or you have product market fit, what are the signals that you've looked for previously?

[James Everingham Steven]

I'm trying to think back at the companies that had strong product market fit, you know, at the highest level, I mean it's not very insightful but like you're trying to hold the wheels on the cart at some point. I can give you an example inside Meta. When this had product market Fit internally and to that is that a product that we built it sort of went viral internally and we had architected it as an experiment, you know and startups often architect things to not be scalable. We did that internally. So you needed an entire developer server for every instance of what we built. It was working well and then Suddenly we had 40,000 developers wanting two and three instances. So like last year I was running around trying to find a spare 150,000 servers in Meta infrastructure to enable this for developers and they were complaining that they couldn't get it. So to me that's a good signal of product market fit. Now a non obvious one. If you're trying to figure out if you're delivering value is like when you take it away, people complain. So like not something I would recommend doing as your go to market strategy. But like that was also when we would take things down internally. We learned what really worked because of people would complain if it went away. I think that was better signal if the tool is working rather than trying to measure lines of code or diffs.

[Global Talent Company Announcer]

This show is brought to you by the global Talent company, a marketing leader's best friend. In these times of budget cuts and efficient growth, we help marketing leaders find, hire, vet and manage amazing marketing talent for 50 to 70% less than their US and European counterparts. To book a free consultation, visit globaltalent.co.

[Podcast Host]

what else when we zoom out. If you just think about your general philosophy for building companies for building products, like how would those who have worked with you say like or describe your philosophy?

[James Everingham Steven]

I remember a long time ago I saw a talk somewhere where people were saying that when you're building a community, for example, you don't want to try to put all of the rigid rules in at first and tell people how to come into your community. And I think of product that way. Like you can think about product and if it's something new, you almost assuredly got it wrong. So my philosophy is to put stuff out there that are more tool oriented, put it out there and see how people use it and then sort of build towards what the real use cases are. And often they're not what you think they are. So that would be one thing and another way of describing it or a way that I would say it is. My philosophy is that in a big company you have to like get stuff right and then before you get it out, when you're doing a small company and you're starting even like a medium company, you got to get stuff out and see if it works before you Put all the energy into getting it right. So design to 50% debug into existence. Relentlessly measure how your users are using the software. Don't fall in love with your idea because of, you know, customers may be using it in a different way and it may be a much bigger opportunity if you actually look at how they're using it, and it's not exactly how you expect it.

[Podcast Host]

When you think about your time, where you're spending time, what does that look like right now? I know there's no such thing as an average day, so maybe what are the priorities? What's top of mind for you?

[James Everingham Steven]

Yeah, sure. For me, it's like, look, one of the reasons I like building companies myself is like, working with great talent and working with people who are just phenomenal. So a lot of my time right now is team building and getting the right team in place. So hiring, getting the best people in the industry. That's part of it. Education is a part of it. You know, this is a new part. Like, what we're building is something very new. And like, everyone is trying to figure out how this stuff works and what its value is. So educating customers, talking with customers, understanding their problems, deeply leaning into the product, like taking these customer conversations and turning them into, like, the product roadmap and what changes we need to make this work for customers. That is a big part of my day right now.

[Podcast Host]

How would you describe the culture there and would you say that's a similar culture that you've built at previous companies?

[James Everingham Steven]

Yeah, I mean, that's one of my favorite things is like, all of the companies have their different culture. I think culture sort of builds on its own and the way to control it one is like, know what you value. And as a leader, you should be very public about what your values are. So I like building teams where we're all kind and collaborative and value the same things. Come from different backgrounds and perspectives, but kind of value those same things very. Bottoms up. When you work with really great people, you know, we try to focus on outcomes and let the team sort of go and invent and explore getting the team as close to the customers as possible. That, you know, customer obsession is like something that I love doing is really, if you're going to build something valuable, you really need to get close to the customer. And I think the other thing that I think is important is that always think if you're building, you'll build the best software if you can put it between you and your customer. And what I mean by that is building Great software is an act of empathy. So like, if you feel the pain yourself, you're going to come up with a better solution. So building things for yourself and using your software and not just beyond, just like customers said this. So I'm going to go build this feature. It's like, use it, get in, figure out ways for your team to like, find value and use it. Even if it's not a direct solution for what you're trying to solve, you'll build better software that way.

[Podcast Host]

Let's imagine you're, you're sitting down with just an amazing candidate. You want to bring them into the organization. How do you describe the vision that you're building towards?

[James Everingham Steven]

Yeah, absolutely. So, you know, I think the long term vision, and I don't even know what long term means anymore, but it might be 18 months or 12 months, is like we firmly believe, just from our experience that agents are going to be prevalent in the infrastructure of the companies and there's going to be hundreds or thousands of agents running ambiently in the infrastructure. Right. So we saw this happening at Meta, we believe this is going to happen with other companies. We're starting to see it. So right now, companies have a lot of internal Web Applications and SaaS Software. We think that's still going to exist, but it's all going to be optimized by a layer of agents that are helping optimize and remove the constraints in between those tools. And if you can envision that world, then it's pretty obvious you need a new type of, you know, agent, operating system or infrastructure layer to manage and control that, because it's very different than regular software.

[Podcast Host]

Final question for you, for those who are listening in, who are excited about this vision and would want to join, where should we send people? Where should they go?

[James Everingham Steven]

Guild AI. We're also in open beta. You can get in and try the product. But we're hiring, we're looking for great people and it's all there on our site.

[Podcast Host]

Amazing. James, thanks so much.

[James Everingham Steven]

Yeah, thank you.

[Frontlines IO Host]

Well, that's all for today's episode of Builders, brought to you by the Frontlines. If you want more amazing content like this, visit Frontlines IO where you'll find the library of more than 1500 interviews with founders, marketers and other GTM leaders, where we unpack the tactical lessons from their journey. And of course, as always, if you do want to launch your own podcast, we'd love to have a conversation with you. Visit Frontlines IO podcast as a service. Mention that you listen, mention you love the show. And we'll give you a 10% discount.

[Podcast Host]

Thanks for listening. We'll catch you on the next episode.