Still a surprising amount of pushback on any AI tooling at all. I think a lot of companies are just about getting there with sort of Copilot. Like, should every company kind of be refounding themselves right now? All those things we believed last year or two years ago, they're still true now because I think in a lot of cases, they're not.

I'm joined today by Chris and Matt from Ona. Gitpod was successful doing millions of dollars of revenue, 60 people, and they kind of completely refounded the company, bought a freeletter.com domain for it. So we're gonna talk about how you managed to keep a 60 person org so nimble, how you go through the rebrand, and then talk about AI agents at enterprises, is something that they're really focused on.

We've naturally gravitated towards a bunch of phrases that have ona in it, that sort of have become synonyms for doing something. One is ask ona. So this naturally came about, you know, it's not some design phrase. Mhmm. We naturally started to go, I'm gonna ask ona, or hey, just ask ona.

Yeah. And it's become a synonym for go and figure it out or have something else figure it out. And the other is Ona's on it. And it it's the same thing. It's like, okay.

Stuff is getting done. Yeah. Ona's on it. Yeah. You're fine.

Right? If we got those two as synonyms for owner, I think we'd be golden.

Yeah. For me, I I really want, hopefully, like, people to take away from that and also this podcast is, like especially companies that aren't allowed to use AI tools for various, like, risk appetite reasons. Like, they can we can actually make that happen, which is where we we're having a lot of success with, like, some of the large banks where regulation isn't completely lined out on some of these things yet. It's not entirely clear what tools they can and should use, and risk appetite is very low. Like, they don't want to egress any of their source code or secrets Mhmm.

Anywhere. Right? So being able to have, like, an agentic experience inside your own VPC, inside your AWS cloud is is super powerful. We have conversations, and they're like, oh, I just assumed we couldn't have this. I'm like, no.

You can. Like, we we can do that. And so, getting that message out there is, like, something that we're gonna be working hard to do for part of the reason we're doing podcasts and things is, we wanna get that message out, you know?

Yeah. And yeah. So I guess the big question here. Right? So you guys are owner, know, f k a a formerly known as Gitpod.

And Gitpod is like was an I'd, you know, heard a lot of good things. Been around for I think five six years.

Yeah. Five years about.

You know, this is like you're doing well. Mhmm. So but you have rebranded. It feels like a pivot. I I suspect you're about to correct me on this.

Let's say you've just pivoted. That's a that's a hard decision, you know. Like, you got 60 people. That's not easy. Could you tell us about why?

It's definitely not an easy decision. And one that sort of took time to to grow. Let's put it this way. So we didn't just wake up one day and go like, oh, let's rebrand. The the reason we'd we're doing that is to sort of reset the the scene, reset the stage.

So git pod, if you take the name and split it right in the middle, it's git and pod. Let's take the pod first. We famously left Kubernetes. We're no longer running on Kubernetes, which is something that we started with. And so the pod in the name simply is technically inaccurate.

And then Git is very limiting to, well, just Git. And in reality, Gitpod or ONA is is a lot more than that. The other thing is it means we no longer need to have conversations about whether to spell the p capital or not. And we're no longer confused being confused with with GitHub. But in all honesty, really what's happening is the the shift in how we write software is so fundamental that this extension, and we can talk at length about we have to talk at length about also what that extension is, but the trajectory that we're seeing ONA take, and these two aforementioned reasons really were enough of a motivation to say, okay, we need a new name, we need to make it clear that this it's not something new, but we're entering a new era of how software is written, and this is a good place to resolve the capital p problem.

To just kind of extend on that a little bit on why it's not a pivot is because everything that Gitpod was doing will continue. Like, we still have this foundational platform where we can deploy developer workloads into, you know, enterprise VPCs. Like, we also have now a cloud offering where you can use it too. And all of the agentic stuff really naturally goes on top of that. Like, we're starting to bake it, like, more into sort of the the primitives of of what GetPod was and what owner will be.

But it's not like we rebuilt all of the technology. Like, this is just like a real natural extension of what we were already doing. The more you spend time with AI agents, like, you become very wary of them. Right? It's like, did that just delete a file that I wanted to delete?

Like, did it just make a change that I'm not comfortable with? Like, what what is really happening here? Like, where's my where's my appetite for Yeah. Guardrails? And it turns out, like, cloud development environments are really, really great primitives to put agents inside.

Because you can kind of think of it as like a little box, a little prison. Right? Can within there, it's fairly safe because it's ephemeral. If we don't like it, we just blow it away. Mhmm.

Right? And so that's really resonating with with customers and people who were previously Gitpod customers as they move towards being owner customers too, we that can build on top of what we already had, enhance the capabilities, and give them agentic story they didn't think they could have because they don't have the risk appetite to use something that runs on a developer's laptop like Clockcode or even Cursor in some situations.

Scaling DevTools is sponsored by WorkOS. WorkOS helps you get enterprise ready. That means they give you all the things that you need to start working with enterprises. Things like audit trails, skin provisioning, role based access control, and single sign on. Let's hear from Utpal from digger.dev,

a dev tool using WorkOS. We haven't had to think about OTH at all. I think support is great. We have a Slack connect channel with them. Issues, if any, there haven't been many, but anytime there have been issues, it's been addressed super quickly.

So odd trail, SSO, stuff like that, we don't think about that anymore. Generally, we don't think about that anymore. And you don't necessarily think about these enterprise features, but they still lead revenue, and it kinda is a no brainer in that sense.

WorkOS helps your dev tools start selling to enterprises much faster. And they're trusted by dev tools like Cursor, Val, and Vercel. If you use their user management auth, you can get your first million monthly active users completely free. Yeah. Could could you maybe just explain what Gitpod was and what owner is now?

So the the way I used to describe Gitpod is it takes a developer four to five hours per week to set up and maintain that dev environment. We see that in studies and our own data. We reduce that to ten seconds. And the way we do that is we take that typically outdated read me and turn it into something that's machine executable. So that if you need to spin up a dev environment, you can do that at the click of a button.

Fundamentally decoupling compute from automation. So it doesn't have to run on your laptop, it can also run-in the cloud. And it's the automation piece that's really key that gives that lets you sort of instantiate multiple of these deaf environments. One sort of very graphic mental model is, like, think of leptops out of a tissue box. You know, you wanna work on something, you're pulling a leptop out of a tissue box, and it's perfectly set up for what you need to do, except without the environmental disaster that a laptop out of a tissue box would be.

So you can get development environments very, very quickly, and that's extremely helpful for humans, and it turns out it's existential for agents. Because to Matt's point, if I wanted to run an agent locally, I would need to give it permission for everything it wants to do. You know, I don't want this to go, oh, I need to reset state. Let me format let me reset the hard drive. Yeah.

That's not what I want to happen. Also, don't necessarily want to give this thing arbitrary permissions when it runs next to my production access or my email. And cloud development environments are really helpful in that. They provide the sort of natural space for agents to run-in. So the way we give agents more autonomy is by giving them a secure space in which they can run, and the sort of instantiate the ability to instantiate many of these dev environments now also means I can have many of these agents running in parallel.

Mhmm. And that's the real power of it. Because as agents get more and more autonomous, we'll need to turn this autonomy into productivity. And the way we do that is by doing multiple things at once. If I now need to sit there and sort of babysit my agents Yep.

I'm not winning that much. Right? Like, I just exchange one activity for another, but fundamentally, it's my own attention that's really expensive. So if I can now give this agent its own space, let it run, let it run, and not need to worry about it doing something that I don't want it to do Mhmm. I can have multiple of these running without my laptop taking off because the fans go nuts.

Yep. And also always in the perfectly set up environment. No no faffing with git work tree or Mhmm. Or I need a Python version of that, and a different Python version here, a Go version there. It's just always perfectly set up and agents and humans benefit from that.

One of the framings I always really liked for for Gitpod was take the best engineer in the company's laptop and the setup and give it to everybody. Now everybody could have that setup on day one. So the way that shows up is, you know, that time to first commit, like, from an engineer joining the company to putting something in production can be can be minutes because they don't have to worry about setting up the laptop. You open owner and you you go to a toggle and you toggle it right, and it starts spinning up an environment for you in the cloud. And the same thing that humans benefit from, like, even in our own company, we, you know, we we build on top of our own platform.

We have engineers join and are productive, like, on day one. And so then the thesis is, well, everything that makes humans more productive actually makes agents more productive too. Mhmm. Like, having a ready to go environment where it's predictable, it's, you know, codifiable, and we can improve it for both humans and agents at the same time is, like, really powerful. The same controls that large enterprises have always wanted to have over sort of the workflows for engineers.

It's exactly the same as you want for agents too. And so it's just like a really natural extension of what was the GitHub platform into what will be the owner platform. Or what is the owner platform, I should say.

Mhmm. And how how have people been receiving it?

So I'd say that there are three groups of of people that are worth considering. One is internally within within the organization, within within owner, and that's come quite natural. Mhmm. So for a long time, we had a a cursor mandate followed by a Cloud Code mandate, and we didn't need to mandate ONA. We naturally saw a transition into ONA.

People just stopped using Cloud Code and started using ONA. And so that reception has been has come quite natural because we saw, you know, how these things naturally built on each other. For our customers and prospects, it's very similar to a lot of them are asked to or see the need to have an AI story to get into this space. They very clearly see how organizations that employ agents are becoming more and more efficient, and they themselves need to do that in order to to keep up or even lead the group. And with that, ONA is very helpful because if you have it deployed already, you know, you have all these primitives in place, you have all these foundations in place already, and we now can give you an agent that works really, really well that builds right on top of these primitives that you may have been using for a very long time already.

And I think like we got onto a little bit before. But I think you've talked a lot about how change is something that you've leaned into. And it's getting 60 people to kind of get on board with something that's, you know, if if not a pivot, a very strong change of focus I guess. Mhmm. How have you been able to do that?

So I think fundamentally, focus has remained the same, actually. Like, our focus is to provide great developer experience, to make developers more effective, to give them great tools, and to make their lives better. A lot of organizations who buy ONA replace their VDI solution with ONA. And anyone who's ever written a line of code in VDI knows it's not exactly an enjoyable experience. So our mission has always

Sorry. VD VDI is

Virtual Desktop Interface.

Virtual Desktop Interface.

So, you know, Citrix or RDP or something like that. Mhmm. A lot of lag, typically under spec machines, locked down, not built for development.

Wow. Okay. So and people what what are the main reasons that people do this actually? Just as a sidetrack. Yeah.

I think this is something that I've come across. I don't know. Like

It's control fundamentally. It's fundamentally the need for specifically large and regulated industries to have control over whether development happens, what happens to the source code, what libraries someone can pull. And ONA with ONA Guardrails provides the same amount of control, except in a much more effective and efficient package that also doesn't make the life of your developers hell.

Interesting. So I guess to the banks, this is how a lot of devs at banks would build.

This is yeah. We we've seen media eyes in a lot of places, and sometimes there's also a better together story. So there can be a world where it makes sense to run ONA and then access that through a VDI. It's still a lot better of an experience because you don't need to do the heavy lifting in the VDI, but you can do that in the cloud that actually has the horsepower to do it. So this is where where a lot of folks have come from.

And so fundamentally, the mission that we're on hasn't changed.

Mhmm. We You're helping, like, the same people with

Exactly. Yeah. Exactly. We're helping the same people in in a very similar way. Except that now, not only do we make their lives better, we also give them superpowers.

Yeah. Because we're multiplying what they can get done in a day.

Yeah. And I guess in a way, it's just like software engineering changed. People didn't use agents before. Now they use them.

Now they use them.

Therefore, you should also offer that. Exactly. That makes sense.

Yeah. It's it's a really interesting thing to talk about as a whole because I I've seen a lot of discussion about, like, should every company kind of be refounding themselves right now? Mhmm. And I think it's a pretty compelling argument that there should at least be a moment to pause to think about all those things we believed last year or two years ago. They're still true now.

Because I think in a lot cases, they're not. Yeah. And you can see this in even large enterprises. Like, if you look at Salesforce, like, they're they're really changing. Say they're changing direction, but they're changing the way they work, and they're very clear about that outwardly, about how much they're embracing AI internally for their own productivity.

You see you see sort of small startups now. They're always trying to figure out how they can enrich their product with AI. Mhmm. And I think the the winners in this space will be the ones who do enrich their product with AI, not just, like, shove a thing on the side. Right?

You end up in a situation where you kind of add AI and it doesn't really add extra value to your product, I think that's when you probably haven't reflected internally enough on, like, how does this fundamentally change our business? And so I know even though it seems quite extreme to change the company name, I think if you did, like, a case study on a lot of other companies, how they're operating now versus how they did in, let's say, 2022, 2023, I think nearly all of them would be different. At least the ones that are being really successful. Like, they will have changed the org structure, like, way they operate, like, the tooling they use. And, like, we see that show up in in a lot of conversations we have with with with our customers because if they haven't done it already, they're trying to figure out how to do it.

Mhmm. And, like, ONA is usually part of that journey. Right? It's like we want our engineers to be productive. We want them to have AI tooling because we acknowledge the productivity gains and benefits that can be had.

We were kind of a bit cautious going into this market because our risk appetite isn't aligned with maybe some some some of the smaller companies or startups or people in different ecosystems. So we're trying to explore where what we can do today. Mhmm. And so and that's when we can usually have a really great conversation.

Yeah. Well, one question I wanted to ask you guys is like how you see because it I can say from my perspective. So as like a, you know, like I was using I think at the beginning of the year I was can't even remember. But I was probably using like when when did cursor oh, cursor got popular like last summer. Right?

So it's like cursor had like kind of a rain. And then I kind of started using Cloud Code a lot. And recently I started to use go back to cursor but use Codecs a lot Mhmm. For kind of just simple things. Send it off to Codecs, creates poor request.

And like how do you see, I guess a people building and how they're how they're using these tools? You mentioned being able to do a lot of things. And then b, like, how do you see kind of because I guess you are to some extent going up against, know, the open AIs of the world and anthropics of the world.

So the way we reason about that is essentially in as a mental model in three waves. The first wave is the way we've written software for decades, where every line was essentially artisanally handcrafted, save for autocomplete and cogeneration. Yeah. The second wave, started about two, three years ago, Copilot, then Cursor, Windsurf, is that tap tap autocomplete. Mhmm.

It didn't fundamentally change the pattern. If an engineer closed their laptop, work ceased to happen. But while the laptop was open, work happened quicker because, you know, you got got that autocomplete. And the thing that changed with agents very fundamentally is that the pattern changed because an agent, as long as it doesn't run on your laptop, once you close the lid, work still continues to happen.

Yeah.

And it's a very fundamental change in how software is written in that it's no longer the human that is very directly producing code, but it's the machine that's doing it. Yeah. So we're really entering sort of this era of automation. It fundamentally changes the pattern. The interesting thing about agents is that agents themselves are a reasonably thin wrapper.

I think Swix is quoted with software engineers perceive agents as an LM in a for loop. Yep. Which is approximately right. I don't think it's doing it full justice Yeah. But it's not that far away.

You know, fundamentally, it's a set of tools, the calling of the LLM, handling all the quirks of that, and then the environment in which it runs. And the former two, the tools and the quirks of the LLM, like, you know, Codex is even open source. Like, there's a bunch of agents that that do that and that do that well. The real magic happens in the model, and the models are commodity. Yep.

So the only thing that's left to differentiate on is the environment, is the integration, and that's what we're going after.

Yeah. I do you well, okay. Model commodity.

Yeah. But debate. Everyone has access to it. Everyone can get an anthropic AI and it's it's your level playing field.

True. So you're if you're competing with OpenAI as of now, they're not giving themselves better access to better models. So that you would be you'd both have the same kind of intelligence.

I mean, OpenAI is, to some extent, a time like the, you know, the codex models for for a while weren't accessible through API.

Oh,

really? Like, some of these things do change and shift, but by and large, you know, models, at least the way it stands today, are the great equalizer.

Mhmm. Yeah. And then so sorry. I interrupted you. So the differentiator is the environment.

It's the environment and the integration. So to Matt Matt said that earlier, you know, so many of our customers come to us and are almost surprised that they they too can have this experience that we offer. Mhmm. And it's really the way ONA is deployed and the way it integrates into your controls, into your VPC, into your organization that makes it work for large enterprises. And that's the part that's differentiating, is the fact that we can bring this experience to really big organizations that otherwise struggle to employ those tools Mhmm.

And have a lot of shadow IT because, obviously, engineers, you know, we

want those tools. Course. Use it. Well, they try to.

Exactly. And so, you know, they'll put that code into some cloud and some rack, whatever it is, and all of a sudden, half of your controls fall apart. And, you know, you can try and fight that, or you can make it easy to do the right thing. Yeah. An owner is making it easy to do the right thing.

Interesting. Yeah. It's been it's been a really sort of wild journey watching this play out, hasn't it? And, like, we can all try and be oracles and predict what happens next, and I don't think any of us can. But there was a there was a ton of time where we all we all thought the models was the moat, didn't we?

Like, models are the thing. And then it was that moment. Was it DeepSeek? DeepSeek released. And everyone, like, panicked and paused for a moment.

Mhmm. And when other models remote, because we just have one open source that's, like, incredible, and it's, like, meaningfully cheaper. And then we kinda went into these wars where people are flicking between ClorCode and OpenAI and DeepSeek and whatever comes next. And, like, switching costs are are cheap. Right?

Yeah. Like, most people have subscriptions to both. Like, I have $10 here, $10 there. Like, I'm happy to try everything. Or if not, I'll cancel my Clawd subscription this week, and I'll get an OpenAI one.

And you see this show up on, know, you Twitter, especially if you're active there. Like, the the zeitgeist changes weekly in terms of what people love and how people are feeling about things. And so, again, once you kind of remove the models and peel it back and go, okay. Let's say it's true that OpenAI have a model that's better than everyone else's. Mhmm.

I think we're already at a point where there's there's good enough other models always, and you're getting into, like, very marginal, not point two, not point 5% better. Yeah. I think the latest release of Opus was, like, 1%, 2% on the SWE benchmark better. Like, it was it wasn't the jumps we've been seeing. Yeah.

Maybe we'll continue to see huge jumps. But I also believe if we do, another company will match them. And so if you peel away all of that, all you have left is, like, user experience platform, where

it

runs, business requirements. You get back into sort of very standard traditional product management. Right? What is the users need and how do we solve it? And the models are one of the things that we need to solve some of our customers' problems, but there's all of this risk appetite business requirement, contractual requirement, legal requirements around it that needs to be solved for.

And we're incredibly well placed to solve that, especially if we stick to our guns of being neutral in all of this. Yeah. It's like if you want to use that model, we'll find a way to help you support it in a way that matches your risk appetite. And so I think, for me, that was the reason I I joined ONA all the all the other things I saw in going on in the AI space is because for me, it makes a lot of sense that somebody who provides a great integration experience and just kind of zooms out into more traditional product management of, what is the problem we're seeing and how do we solve it? And how is AI part of that, but not the entire solution?

It is is really, really powerful, I think.

I guess you've kind of seen who buys like I guess the people that were buying GetPod. You know, you understand that user very well. And I guess like it'd be okay if like you haven't got like every startup in SF is using owner. If like because maybe the environment stuff is less important to them. But it it probably seems like it would become a lot more important with like the more complex your organization is.

And, of course, we want every startup in the Okay.

Yes. So

use ONA. And you'd actually benefit from Yeah. From doing that too. You know, it's not just for for the large organizations. They clearly have a lot of organizational and compliance controls that they need to put in place where ONA is immensely helpful.

But even in in my spare I mean, I'm a bit brainwashed here, but even in my spare time, if I I would use owner just because it means I no longer have to as someone recently put a prostitute my laptop for agents, you know, I it means now I can do this from my phone. One story I like to tell is I I have a three months old son now, and I spend a lot of time in the evenings once he's asleep and he likes to fall asleep on me. Yeah.

You can't move.

Exactly. I can't move. I definitely can't have a laptop here, but I can't have my phone. And the way I frame that is, I'm now three times more productive on my phone than I was six months ago on my laptop. And it's not it's a bit hyperbole but not much.

Yeah. I I completely see that actually. Because I I've been doing sending off like, you know, you just describe what you want. And I've been using Codex mobile app like ChatGPT mobile app and like Yeah. It is incredible.

Do you guys have a app, actually? If you've launched, a mobile app?

We haven't launched a mobile app. We we had a big conversation about that, didn't we? And we're trying to figure out what the right thing to do was. But I think we we've started out with the mobile form factor. So if you go to our mobile website, like, everything works.

One of the really nice things about cloud development environments is I see lots of people on on Twitter trying to figure out how to get Cloud Code to run from the phone so they can interact with it with a way. It comes for free with with with ONA because, ultimately, we're spinning up a a box in AWS. And so you're just interacting with that remote box and everything's happening there anyway. So we very naturally just get mobile functionality, which is great. So for now, we're kind of experimenting because one of the really interesting things that we haven't we're still trying to form an opinion on the right way to do it It's like jumping into, like, manual code edits Mhmm.

On a phone. There's a few different apps that have tried a few different ways to do that. We don't love any of them. So we're trying to figure out what our opinion is on that and how we want that to show up. Once we figure that out, then I think we'd be, you know, maybe explore apps as we get bigger.

But, like, I think you can get awful an awfully long way with a, like, a progressive web app and a mobile website these days. So probably stick with this a little bit longer. But once we get into maybe deeper coding experiences on the phone, that might be when we'll, like, actually, let's see if it makes sense to get a little bit closer to what you can do natively.

Yeah. I actually interviewed one of the a guy who has like the most popular git mobile app in the App Store. And I don't know. Like maybe it's gonna actually be like a resurgence of like this is actually like a lot of developer experience happens like on your phone.

I mean, we're we're seeing something very similar when back in the day when we started, we started as a code review platform that gave you an IDE in the web. That's how we very originally started, and then very quickly realized that the actual problem to solve is the dev environment underneath all that. What's interesting now is that with agents, we're seeing a resurgence of having the IDE in the web. So the way ONA is built is really to sort of progressively engage with code. You can go deeper as you need to.

You know, sort of at the first layer is doing multiple things in parallel, so you get a very precise, let's say, overview of what your different tasks what state of your current tasks, and how long will this still take, and really helping you sort of scale your attention across them. And then when you click on one, you get the conversation that is really designed to give you maximum context in in a very brief Mhmm. Glance. You know, you don't have to go and scroll up a whole bunch to understand where things are at. At the end, you also get a summary, and if that's not enough, you can either from within the conversation or in the sidebar, essentially open Versus Code web right there and then, and it's full Versus Code web.

It's not some clone or some stripped down version, but it's actual Versus Code, and the intent here is to make you feel at home. You know, we're so used to Versus Code and also to a full IDE, but they're also quite heavy. Imagine doing five things in parallel and needing to have five Versus Code windows open and trying to understand while you command tap between them which one's which. Yeah. Whereas by putting it in that same interface, you can very easily switch between these conversations, and you're always in the right context, and you can bring Versus Code in whenever you need.

And if you need to go one step deeper, there's a button in the upper right corner that lets you open this in cursor, JetBrains, Versus Code. You can SSH into that box and use VI if that floats your boat. So this ability to sort of engage with the code at the appropriate level is really important, and having the ID in the web really makes so much more sense now than it did five years ago.

It all loops back to to that theme about being, like, sort of the center of the way you work and integrations. Right? Like, I think a lot of the other sort of AI tools change the way you work and sometimes for the better, sometimes for the worse. But, you know, if if you if you don't love working from the terminal, for example, like, that seems to be the zeitgeist at moment. People wanna do that.

Might It not necessarily meet how you like to work. I was was a big JetBrains user. I I was used to spend a lot of time in Goland and then changed to Cursor to use Cursor's AI models, but never really fell in love with Versus Code, especially maybe it's just like muscle memory at this point, you know, I can navigate very quickly around one. And so I think it's very, very powerful to position yourself in the center of however you wanna work will support you and will give you sort of tooling around it to be able to integrate with. Yeah.

You know, if if you are a JetBrains company and you've got, like, tons of licenses, please go ahead. We give you the ability to switch on and off specific IDs at the organization level. So if you are a big company, it's like, I want a pilot cursor. We Like, can enable you to turn it on for a week and then turn it off, and you can gather some data and stuff like that. So very, very powerful to to be in the middle of all this and to be a bit more like, we're not gonna change the way you work.

Like, we wanna work with you the way you wanna work, but here's some extra tooling and and sort of agentic experience that kind of fits into your workflow rather than changes it.

Yeah. I've got so I I I wanna ask like kind of just a general question because I guess you're seeing a lot of different don't know. I felt like we talked about dev Twitter and a lot of it is just like people that, you know, it's like startups. Right? Like a lot of it's like there's not so much, you know, enterprise stuff.

Right? What what kind of how are people adopting it right now? Like, what is the what is the status? Like, are people what's it looking like in in the enterprise space in like AI? That's a great question.

So what I mean, one

of the things that we basically, what you're seeing on on especially, like, dev Twitter, I would say is probably, like, the top 10% of people adopting things. Right? They're probably at the forefront there. They're very much early adopters at the top of the curve. One of the things we're seeing a lot in our conversations is I think a lot of companies are just about getting there with sort of Copilot.

Like, we're starting to think about, like, how Copilot can fit into their workflow. Still a surprising amount of pushback on any AI tooling at all. As in, like, we don't really know how to think about this yet. We don't have to reason about it. We don't know how it's gonna show up in our company.

Like, we're not sure that what people are doing over there applies over here. So I think a large amount of of what we do in some of this is almost, like, educate a little bit too and show what is possible, what's capable, what tools are out there. One of the real challenges we have is there is you know, in every company, there's gonna be those people who are in the in dev Twitter, and they want they wanna use all this crazy, like, literally just came out bleeding edge AI tooling at work. And so, actually, we love it when those people exist, because they can actually help us navigate the conversation and say, like, well, you know, if you wanna do that at home, we can help you bring that to work too. But we can help you convince your boss that, like, this actually makes sense for you and that it it can work within your company.

And it's not just for startups. It's not just for small companies. So something we do have to spend time doing is, like, marrying up these these few things where there's, like, enterprise risk appetite. So, like, what the company wants to be true and then what the engineers want to be true. And then if we can get ourselves, like, in the middle of that conversation, we get do a really good job generally.

But finding the right people to to kind of connect all those things off is is is quite challenging. Mhmm. And so you almost end up kind of helping companies navigate their own org chart a little bit, which is which is really challenging. But it's it's very rewarding when you when you can figure out.

Yeah. I remember I always remember like I did some enterprise sales back in the day and it was always funny when you're like, oh but you guys already have an initiative to do this thing like over and there and they're like, do we like and it's like you just think that they would know what's going on in their company, but sometimes you have a better view on it. It's kinda

Yeah. I think that the theme of of two speeds we find play out at different scales over and over. So one is, you know, you got Dev Twitter, which is really sort of plugged in, forward leaning, sometimes critical of of things that are happening, but generally, really sort of at the at the forefront of that wave. And then you have most engineers out there who are not Dev Twitter, who still believe Copilot is the pinnacle of AI tooling right now. And the same plays out within large organizations.

I was really surprised to see that, specifically, enterprise organizations really understand that AI is a fundamental game changer. Like, at the leadership level, the moment you get into boardrooms and c level, the conversation isn't, do we need AI? The conversation is, how do we get AI in here quicker? That's the conversation.

So they do see it.

They do see it. And they also see the potential that that has, and they also see the risk that it has. Like, it's really difficult to sort of outperform a Fortune 500 company, but they also see the innovator's dilemma and that they're big and potentially slower than others. And AI is this crazy accelerator that now, you know, makes it possible for a small team to take a serious chunk out of someone's business. Yeah.

And they see that. And at the same time, they're facing exactly the situation where a lot of the engineers don't yet see it and where their own compliance requirements might make it hard to act on this desire to roll out AI based dev tooling, specifically agents. And that's really where we can be very, very helpful because we can bring these controls, and we can bring this efficiency that that folks seek. And we've seen this play out with several of our customers where there was excitement in the forward leaning group that the Matt just spoke about. And there was excitement at the c level where they saw that we need to bring this tooling in, and we help connect the two and bring them together in a way that actually works for these organizations.

Chris and I will spend quite a lot of time with our customers, like both engineers and c level to do this. And I think that's that's a pattern I think we'll see more. I think you will see engineers being part of deal cycles and go to market because it makes sense. Like, you are like, how how often do we hear the like, we wanna speak engineer to engineer. And and they wanna have a conversation with, like, there's a lot of hype in this space.

I wanna cut through it. Can I speak to someone who is using it? Like, can you know, I wanna speak in a bit more technical level what they're seeing, what they're feeling. And, like, I think that's something you should lean into is is, you know, those engineers in in your company who are inclined and are power users, like, let them speak to your customers. You know?

That that passion will shine through, like, a lot. And I think that's really powerful.

I'm hearing, like I don't know if it's just, like, random, but it feels like people are talking a lot about, like, forward deployed engineers a

lot more than they were like Absolutely. Mhmm. We're seeing that too.

Yeah. We do.

Which I think makes sense because I am I mean, the the the sort of the the the biggest thing I read about this was, OpenAI, we're hiring a a whole bunch of forward deployed engineers specifically,

which Yeah.

Again, makes sense. Like, they're talking about joining companies and, like, fine tuning models and, like, really helping them deploy AI. But I think if you zoom out, like, one step back and think about what you're kind of putting an agent in to do and hopefully scaling out, so you have lots of agents doing work, is you're kind of fundamentally changing the way you work in your business. Right? And so there's gonna be lots of, controls you need to put in place.

There's gonna be lots of advice you wanna seek very quickly that's very scoped to your organization. And so having someone who's if you've committed to purchasing a product or, like, an agentic tool, and you just wanna you wanna make sure it's successful. Mhmm. So one thing we talk about often is nobody buys, like, a dev tool, especially in a large company, and not wanting it to be successful.

Mhmm.

But part part of our role is to make sure you are. Yeah. It's not just like, oh, we sold you it. You've signed the deal. Like, we'll see you in twelve months.

Like, it's no. We like, we need this together. And, like, you're in, we're in. Let's figure out how to do that. And I think forward deployed engineers as a title is quite an interesting way to, explore how you can make that possible.

It's like, we'll take someone who would maybe traditionally be in our engineering team. Yeah. Like, we'll put them inside your company, especially for the larger contracts. And their entire job is to help you be successful. Yeah.

Like, if you're missing a feature, we'll figure out how to scope it out for you and maybe build it out. Yeah. If you've got some specific questions about how you can use it and best practices, like, we'll plug you in to our other customers and help you figure that out too. And so I think I think it's really compelling, and I I think we'll continue to see more and more of this.

Yeah. Do they this just I actually hadn't don't know much about the role. Is it like, they're typically, like, forward deployed engineer will just be, like, if you start working with, an a Fortune 500 company, they'll basically be there, like, seven to five five days a week.

I think the beauty of the role at the moment is it's kind of up for grabs a little bit depending on the company. You can kinda shape it a little bit into what makes for for your company. But I think in incredibly large enterprises, yeah, that would make a lot of sense. At least for a period of time, they would spend four or five days a week with the company as if an extension of their engineering team would be the bridge. I think as hopefully, as that relationship matures and gets stronger and, the understanding of the products exist within the company, you might find they can be on two accounts perhaps.

But, generally, yeah, it's an extension of their engineering team that bridges into the companies. Yeah.

A lot of what that role needs to do is really navigate the organization. Mhmm. So we call them forward deployed engineers, and there's a lot of engineering that needs to happen. But there's also an equal amount of understanding the needs and desires and fears, to some extent, that exist within that organization that they're becoming a part of. And there are fewer things that work better than having them batched in and be part of the coffee kitchen conversation.

Yeah. So it's incredibly important that they really become embedded in that organization, get as close as they can and as as compliance permits so that they can effectively help that that customer and navigate that change that is happening.

Amazing. Are you guys hiring forward deployed engineers if anyone's listening?

We are hiring forward deployed engineers.

Nice. We do. I think that's all we got time for. So I just wanted to see see if there was anything that you wanted to point people towards. And you've got a nice.com domain now.

That's right. 3letter.com domain. So head over to owner.com and sign up, try it out. Very excited to see how this works for you.

Amazing. Well, thank you Chris and Matt for coming on and thanks everyone for listening.

Pleasure.