Swapnil Bhartiya: As we all know that AI is speeding up infrastructure work, but it is also speeding up mistakes. When AI can generate terraform or pulumi in seconds, the real question becomes who is governing what gets deployed When AI can. AI is speeding up infrastructure work, but is also speeding up mistakes. And when AI can generate terraform or pulumi in seconds, the real question becomes who is governing what gets deployed and how fast can teams catch problems before they hit production? SpaceLift is aware of this problem and they are building a control plane for infrastructure as code, including new natural language capabilities like SpaceLift Intelligence and SpaceLift Intent. And today I have two guests from SpaceLift. John Henry Archer, SVP of Global Sales and Jonah Co, SVP of Product and Design. Jonah. John Henry, it’s great to have you folks on the show.

Jonah Kowall: Thank you.

John Henry Archer: Thanks for having us. Thank you. Thank you for having us.

Swapnil Bhartiya: It is my pleasure. Of course, we have been covering spacelift here on a regular basis, but since you folks joining a new role, so I would love to hear from you. Jonah, can you just tell our audience what is spacelift all about and what problem was it built to solve?

John Henry Archer: Our platform is specifically designed to help teams deploy various infrastructure as code technologies in a safe, repeatable way into their production environments. And this enables teams to move quickly while having the guardrails and the governance that’s required to avoid major production issues and outages. Beyond that, we also have capabilities to unlock AI use cases in various ways that we can discuss and we continue to expand the platform capabilities. We’re also very tightly coupled and work specifically on OpenTofu, which is another important technology that’s part of our platform.

Swapnil Bhartiya: Excellent. Thank you. And if I’m not around, you folks recently joined spacelift. So I would like to know a bit about your own background, what you bring to SpaceLift and this ecosystem, and even more importantly, why SpaceLift? So it’s twofold question and John, we can start with you.

Jonah Kowall: So I’ve been in tech sales for over 20 years. I’ve been a part of building go to market teams at hypergrowth companies like Pure Storage, Lace Work and Own Backup. I love building high performance go to market teams. What attracted me to spacelift was the executive team and their vision. I really believe that the infrastructure automation market is being recalibrated right now and I believe that SpaceLift is positioned to define the next generation.

Swapnil Bhartiya: Jonah, from your perspective.

John Henry Archer: Yeah. So from my side, I spent 17 years as a practitioner. I spent about four years as an analyst at Gartner, and for the last decade or so I’ve been building products mostly in observability, but also in data infrastructure at companies like AppDynamics and Aiven. I’ve been both a CTO and also a head of product and design. The thing that interested me with spacelift is not only the way that we’re able to meet so many different customer deployment models, but in the age of AI and the velocity of code, the infrastructure has to keep pace and we enable organizations to do that safely. I also think it’s critical to be underpinned by an important open source project. And so for me that really attracts me because I’m also a maintainer of another CNCF project called Jaeger in the tracing area. And I also work on OpenSearch, which is another Linux foundation project. So I really have a passion for open source work and SpaceLift is aligned with all of that.

Swapnil Bhartiya: SpaceLift just hosted IC Conf in May under the theme Keeping page. Can you talk about the theme? Can you talk about what kind of concerns worries practitioners have when they showed up at the event? What kind of questions, what kind of discussions you’re hearing there?

Jonah Kowall: Firstly, I want to thank the entire team at SpaceLift who helped us put on that awesome event. It’s really great for the community and exciting to see it continue to grow. And what I keep hearing from enterprise customers that mirrors the IC comp data was that they’re not asking whether to use AI for infrastructure. That decision is already been made. They’re just asking how to do it without introducing governance failures, audit gaps or security incidents. The stat that stood out the most to me was that 89% of organizations are planning to adopt Agentic AI for infrastructure. 24% of those are planning to do it within six months, which represents a huge pipeline opportunity for SpaceLift. The urgency is real and we’re uniquely positioned to help there.

John Henry Archer: So I think JH summed it up very nicely. But the one thing I’d like to add is that when we poll the audience, IaCConf and I was able to attend, most of them are sick of AI. They hear about it too much. It’s like overwhelming. But then when we ask them what they want and we sent out a survey, then most of them say they want to hear about AI. So it’s sort of this weird thing where the infrastructure teams are very cautious about how they introduce something that could potentially hallucinate and cause problems, but do so safely. So I think that teams are a bit scared to take their hands off the wheel, so to speak. But at the same time their teams are pushing them in that direction. Developers want to get these things into production and that’s the challenge for platform teams today.

Swapnil Bhartiya: Did you say that most of the people did not like AI? So what is the reason? Because I feel that today everybody loves. I mean, there are a lot of segment. They don’t like it. And the reasons are different. There is no one set of reason that everybody. Some. There’s a lot of FUD going on then, of course.

John Henry Archer: So

Swapnil Bhartiya: what was their biggest concern? Where you also feel that, hey, this is a space where spacelift open Internet can enter and solve that problem.

John Henry Archer: Yeah. So I think from the poll it was more that they’re sick of hearing about it because every company, every organization, it’s top of mind. I mean, when I look in our development channel, as you know, the new model from Claude just came out and everyone’s trying to do interesting things with it, experimenting with it, learning what it can and can’t do. So, you know, all teams are going through the same thing where they’re trying to use these tools to increase velocity and find. We found new interesting bugs in our software that have probably been there for years. Really difficult solutions are being able to be solved with the help of AI. So it’s a tool that everyone is constantly using, but yet we’re bombarded externally. That’s all we hear about. When you go to a conference or you listen to someone speak, it’s. It’s sometimes a bit too much and that’s kind of the feedback. But then people want to hear about it too. So it’s, it’s a conundrum.

Swapnil Bhartiya: It is. The thing is that AI is not going to go away no matter how it’s going. And here, you know, our show, Agenting Enterprise, we, we talk to practitioners, not in the hype of the AI, but how AI is in production. And one of the concerns that we hear a lot as AI is no longer a chatbot, it’s no longer telling you things, it’s actually executing things, it’s writing code, deploying code. So, John, if I may ask you also, if I’m not wrong, one of the themes of ISE conf was also governing AI generated infrastructure code. With AI now writing and deploying ISE at scale, how is SpaceLift helping or acting as a control plane for those who are using infrastructure code? What are you folks doing in this space to help with governance? Because that is becoming a really, really important topic these days.

Jonah Kowall: Yeah, it’s a, it’s a great question. So I think like our enterprise customers and prospects that we’re working with, they’re not looking for for like a lot of the AI hype, right. They’re looking for a platform that lets them capture productivity gains of AI assisted development without transferring risk to production infrastructure. The enterprise customers that we’re working with aren’t just looking for a faster deployment tool, they’re really looking for, they’re working across multi cloud, multiple services within each cloud with many cloud accounts. Right. And so when they think about governance at scale, it requires a system of record enforced policy and auditability. And also faster deployments are nice to have as well. I believe the IaCConf community, right, the platform engineers, the SREs, the Devop practitioners represent both the technical buyer and the internal champion for us in many of our enterprise deals. And I think the fact that SpaceLift continues to contribute to this community is absolutely a strategic differentiator for us.

Swapnil Bhartiya: Jonah, can we talk about, of course, spacelift Intelligence and spacelift Intent. What do these two capabilities do? And also talk about is it going to complement or replace the approach people already have with infrastructure as code? Because that could also become a big challenge. You know that if it is complementing, it’s totally different. If it is replacing, then you have to think too much about it.

John Henry Archer: So let’s talk about that as well, for sure. So spacelift Intelligence covers like a few different use cases. So I’ll kind of like walk through the progression that a customer typically goes through. So the first part is what we call Infrastructure Assistant and this allows you to query your environment. So for example, I might ask questions about what I have deployed, I might ask questions about my policies, who has access to certain things, even configuration. So that’s read only. There’s no risk, nothing gets written, nothing gets changed. That’s typically how customers will start. And then intent allows you to actually define infrastructure as code. But unlike an LLM which writes call it IAC code in whatever language you’re using, whether it’s Terraform language or pulumi or someone else, we can do that as well. But intent actually creates the policy in a different type of language, essentially, which is translatable to multiple IAC technologies. So it allows us to do a lot of different things because we’re representing the structure of what you intended to deploy or do. And that could be a policy, that could be an actual template itself. There’s lots of different use cases for writing. The other piece of intelligence, which is the last piece, is an MCP server that interfaces with both of these. So if you. On your desktop, let’s say we’re using Codex or Claude, we have an MCP server we host that has the guardrails around it so that you can use that in your environment safely and you can prevent issues from occurring. So many customers will use either the technologies in the product and they’ll use the MCP for other use cases. So we unlock all of that for them.

Swapnil Bhartiya: Specifically, when it comes to AI, AI infrastructure. And then when we talk about infrastructure as code, is it ditto? Exactly. Same as we talk about the. We cannot use the word traditional infrastructure codes or when we talk about AI workloads, a lot of things change or it’s no, it’s the same infrastructure, nothing changes. Can you talk about that aspect?

John Henry Archer: Yeah. So I mean, the infrastructure does change, but it doesn’t change materially from an infrastructure as code perspective. So whether we’re deploying on an instance that has a GPU or we’re deploying on an instance that has traditional CPU workloads on it, it’s basically the same. Because the way that you have to sort of think about it is the platform team that’s using infrastructure as code, they lay down the foundation for the application which runs on top of it. So if you were running an application that was training a model, for example, on top of a gpu, the infrastructure as code would set up the operating system and the other components on top of that GPU instance, you might implement a kubernetes deployment on top of that. And then the application teams would typically deploy the applications on top of that infrastructure as code. So we’re really there to orchestrate the networking, the storage, the compute, all of the other pieces together and then the application uses those things. So we don’t really do something different, whether it’s a GPU workload or a CPU workload.

Swapnil Bhartiya: And John, when you talk to enterprise customers, can you talk about from business side, what kind of concern, what kind of concern that keep coming up, what are the gaps that they identify that they are trying to close? And that’s where you say, hey, this is what spacelift Intelligence is here to solve.

Jonah Kowall: So I think if we’re kind of going to the. In general, what I keep on hearing from large enterprise customers, especially at the CIO CTO level, is that there’s absolutely the development teams are running faster than they ever have before. There’s always been friction between the infra teams and the development teams. But what we’re seeing is with AI helping speed up development, that friction is continuing to increase and so we’re in a unique situation where we’re having conversations with them about helping infrastructure teams keep pace with the development teams.

Swapnil Bhartiya: Since you folks work very closely in the open source space, there are two or three things happening in this time. First of all, CRA is coming in Europe where if you’re shipping any open source, the responsibility is on you. I mean company like spacelift, you folks, I’m pretty sure fully prepared for that. So that is not a concern. But two other concerns that I hear from open source folks, and they do complain about AI a lot, is that a lot of code which is contributed, written by AI. There is nothing wrong in that. The problem is that a lot of time people don’t even know what they are doing. So they fully trust AI and submit the code, which becomes a big problem for maintainers. Second is that a lot of projects they have started to use AI to write code. Maintainers have started, which also mean that humans should still dictate the direction of the project, not the AI. But they also have to worry a lot about governance. So can you also talk about giving them the freedom, also ensuring that they don’t get spammed? With a lot of AI generated code they are able to maintain the. I don’t know if we can use the word code hygiene here through right governance practices and also remove bottlenecks so that there’s no rigidity there in whether you can or cannot use. So it’s like very hodgepodge question, but because open source is kind of confusing in that way. So what would be your take? When we look at Terraform, OpenToFu, Pulumi, Cloudformation, Ansible, a lot of tools are there AI is there talk a bit about how are you helping them maintain this balance with governance? Speed, safety, privacy, security, compliance?

John Henry Archer: For sure, Yeah, I can talk to that one definitely. So I can tell you from when I put on my Jaeger hat as a maintainer, we use quite a bit of AI to actually help gate and test quality of pull requests and things that are coming into the project. We also have some pretty stringent AI policies where we ask the person contributing how much AI was used to generate the code and generate the, you know, their submission to the project. And we review everything people do at the end of the day. So it’s down to our bandwidth specifically, but we use a lot of tools. So for example, before we would look at anything, it’s been reviewed by at least two different AI systems. It’s been categorized, it’s been filtered and a lot of that Work, which we used to have to do manually, is now automated. So it actually makes our job easier and it improves the quality of the code as well. Now the challenges on the OpenTofu side, as you may know, this was a fork of a HashiCorp Terraform project which had a more restrictive license. And we’re very careful about making sure that no code leaks into that because of the license. And therefore there’s no use of AI in OpenTofu. We don’t generally allow it because of the licensing challenges that we have. Now the good news is that the we’re building a new engine for OpenTofu which is much more flexible and performant and because that will be entirely new code base, we have to worry less and less about that problem in the future. So as OpenTofu goes in its own direction, we will start to differentiate more and more. And that’s already been happening across the projects because they are compatible for the most part OpenTofu and Terraform, but they’re definitely going in different directions, which is good for the users and that’s the point of it.

Swapnil Bhartiya: John, from your go to market side, what kind of opportunity you see here for spaceleaf or you are personally excited over because you do see those gaps and you do see that spacelift is filling those gaps or will further fill those gaps. What are those?

Jonah Kowall: You know, first of all, I think spacelift Intelligence absolutely marks an inflection point in the industry. The move from a deployment automation tool to an AI native control plane changes the strategic conversation with our enterprise buyers. I believe that we need to continue to build out spacelift needs to continue to build out our go to market functions and processes to support our existing customers and support our large enterprise prospects that we’ve been working with over the last few years. And then I think for us also last but not least, but we need to invest in our partnerships to help drive positive outcomes for every customer and continue to expand our reach.

Swapnil Bhartiya: And from Jonah, from your perspective, from product perspective, what gap do you see and what opportunities do you see there as well?

John Henry Archer: I mean, I think there’s a lot of gaps in the market, whether it’s expanding, what IAC can do. So as I mentioned before, we have the infrastructure and the application, but we also have layers below. So where Terraform and OpenTofu work, there’s plenty of things that are lower level to the hardware that still exist. So from a spacelift perspective, we want to really help our customers solve all of their IAC needs, whether that’s moving up closer to the application or moving down closer to the hardware. So I think that there’s lots of areas for us to expand into that will really solve customer challenges as we continue to build out the platform and add more capabilities and modules on top of what we’ve built today. So that’s what I’m excited for, is really solving those holistically.

Swapnil Bhartiya: Excellent. Thank you. Of course we are talking about these two capabilities and I really now I don’t like to ask this question. Earlier I used to love asking this question, what does your road bed look like? What are the things in the pipeline? Because things change so fast that sometimes you will react so fast so all your plans will be gone. But still I will throw that not about like concrete roadmap, but looking at the market, looking at some of these pain points, looking at these gaps, what are the things in your pipeline that you folks are working on? What does your roadmap look like after this?

Jonah Kowall: This.

John Henry Archer: So we have a few different options that we’re still juggling in terms of the bigger picture. We’ve got a lot of short term things that we’re building to improve the current platform, increase scalability, new deployment models. But the bigger thing that we’re really solving is going back to that issue. How do we help customers instead of just deploying kubernetes and managing it, how do we help them deploy their applications more effectively? That’s one of our big bets that we’re looking into deeply so that we can start to look at that whole stack of infrastructure all the way up to the application. And that’s kind of one of our big goals and objectives for the next six to nine months.

Swapnil Bhartiya: John Haree, Jonah, thank you so much for joining us and of course being part of this conversation with the company. Thank you so much for sharing all these insights, talking about the pain points, where things are heading, how SpaceLift is solving some of these problems in the production and folks who are watching, please go check spacelift, especially when it comes to AI and infrastructure and infrastructure code. And I would love to have you folks back on the show. Thank you.

Jonah Kowall: Thank you so much.

John Henry Archer: Thanks a lot.