As part of KubeCon + CloudNativeCon North America, theCUBE’s Rob Strechay sits down with Marcin Wyszynski, co-founder and chief research and development officer at Spacelift (and OpenTofu cofounder), to unpack how AI is reshaping infrastructure as code and platform engineering. Wyszynski traces Spacelift’s origins from “multiplayer Terraform” needs to the recent announcement of Spacelift Intent – an experimental, open-source-first capability with a commercial, multiplayer layer. He explains how Intent accelerates day-zero prototyping, imports ClickOps-created resources and even breaks up overstuffed Terraform states. The discussion highlights why governance and policy-as-code guardrails matter when putting large language models in infrastructure as code loops.
Viewers get a preview of what to watch this month in Atlanta: open sign-ups for Intent on the Spacelift platform (post-KubeCon), improvements to Spacelift’s Kubernetes integration and Spacelift’s first in-person IaCConf focused on the future of IaC. Wyszynski underscores a pragmatic approach – extending Terraform/OpenTofu rather than replacing them – enabling teams to move quickly without sacrificing oversight. It’s a timely look at how cloud-native foundations, AI assistance and strong governance can reduce toil as AI workloads scale.
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
KubeCon + CloudNativeCon NA 2025. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Register for KubeCon + CloudNativeCon NA 2025
Please fill out the information below. You will receive an email with a verification link confirming your registration. Click the link to automatically sign into the site.
You’re almost there!
We just sent you a verification email. Please click the verification button in the email. Once your email address is verified, you will have full access to all event content for KubeCon + CloudNativeCon NA 2025.
I want my badge and interests to be visible to all attendees.
Checking this box will display your presense on the attendees list, view your profile and allow other attendees to contact you via 1-1 chat. Read the Privacy Policy. At any time, you can choose to disable this preference.
Select your Interests!
add
Upload your photo
Uploading..
OR
Connect via Twitter
Connect via Linkedin
EDIT PASSWORD
Share
Forgot Password
Almost there!
We just sent you a verification email. Please verify your account to gain access to
KubeCon + CloudNativeCon NA 2025. If you don’t think you received an email check your
spam folder.
In order to sign in, enter the email address you used to registered for the event. Once completed, you will receive an email with a verification link. Open the link to automatically sign into the site.
Sign in to gain access to KubeCon + CloudNativeCon NA 2025
Please sign in with LinkedIn to continue to KubeCon + CloudNativeCon NA 2025. Signing in with LinkedIn ensures a professional environment.
Are you sure you want to remove access rights for this user?
Details
Manage Access
email address
Community Invitation
Dimitri Vlachos, Spacelift
As part of KubeCon + CloudNativeCon North America, theCUBE’s Rob Strechay sits down with Marcin Wyszynski, co-founder and chief research and development officer at Spacelift (and OpenTofu cofounder), to unpack how AI is reshaping infrastructure as code and platform engineering. Wyszynski traces Spacelift’s origins from “multiplayer Terraform” needs to the recent announcement of Spacelift Intent – an experimental, open-source-first capability with a commercial, multiplayer layer. He explains how Intent accelerates day-zero prototyping, imports ClickOps-created resources and even breaks up overstuffed Terraform states. The discussion highlights why governance and policy-as-code guardrails matter when putting large language models in infrastructure as code loops.
Viewers get a preview of what to watch this month in Atlanta: open sign-ups for Intent on the Spacelift platform (post-KubeCon), improvements to Spacelift’s Kubernetes integration and Spacelift’s first in-person IaCConf focused on the future of IaC. Wyszynski underscores a pragmatic approach – extending Terraform/OpenTofu rather than replacing them – enabling teams to move quickly without sacrificing oversight. It’s a timely look at how cloud-native foundations, AI assistance and strong governance can reduce toil as AI workloads scale.
In this post-KubeCon + CloudNativeCon NA interview, Spacelift Chief Marketing Officer Dimitri Vlachos joins theCUBE’s Rob Strechay to discuss the evolution of infrastructure-as-code (IaC) and the emerging role of generative AI in DevOps. Vlachos outlines the friction between the speed of "ClickOps" and the rigorous "ceremony" of GitOps, introducing Spacelift’s new "Intent" solution designed to bridge this gap. The discussion highlights how Intent leverages AI to allow developers to deploy infrastructure via natural language while maintaining essential guardra...Read more
exploreKeep Exploring
What is the origin story of Spacelift and how did it come to be?add
What are the advantages and limitations of using ClickOps compared to IaC and GitOps for managing deployments in AWS?add
What are the challenges of using GitOps and Infrastructure as Code (IaC) for developers during rapid deployment for non-production use cases, and what solution has been introduced to address these challenges?add
What are the hopes for Spacelift's progress and developments over the coming year as discussed in the context of an event in Salt Lake?add
>> Hello and welcome to this post- KubeCon + CloudNativeCon 2025 follow up. I'm Rob Strechay, and today we're going to talk about infrastructure-as-code and what happened at KubeCon + CloudNativeCon back in Atlanta. To help me do this, I'm
really excited to be joined by Dimitri Vlachos, who's
the CMO of Spacelift. Welcome on board, Dimitri.
Dimitri Vlachos
>> Great to be here, Rob.
- Again,
Rob Strechay
>> we saw each other a whole bunch, but we were running in
different directions a lot of the week, except on day zero, and we'll get to that in a minute. But can you start by going into the origin
story for Spacelift? Because I think it's really intriguing and important about how this community has
really been built out.
Dimitri Vlachos
>> Sure. So, one of our
co-founders, Marcin, our technical co-founder, head of R&D, after a career at Facebook
and Google, running large SRE and DevOps groups in
various different roles, he was doing some contracting and he was helping a lot of companies and realizing that the way they were managing their infrastructure
with IaC, with Terraform, there had to be a bigger
way when you really want to scale this. And so, he actually
started to build some tools as a contractor to help
organizations with that. And that really led into
the idea of Spacelift and started to productize that and creating a real offering that could help people scale
their infrastructure-as- code operations to both
have the speed they need to deliver infrastructure, but the control they
needed to really be able to have the governance and security behind the scenes of what they needed for infrastructure. So, that's the origin story of Spacelift.
Rob Strechay
>> Yeah, I think again, it's
when you start to look at how people are bringing together
cloud-native applications and infrastructure and really trying to do it
repeatably, having been at one of the hyperscalers and knowing that there's definitely a different levels of skill that people have. And one of the major things that people talk about
is really ClickOps versus infrastructure-as-code. What do you see as the major differences as you're out there talking to customers?
Dimitri Vlachos
>> So I mean, if you step back and say, "Okay, cloud
operations, cloud infrastructure, the original people started
deploying by going into consoles of the different cloud providers. " Really, at the time when this
really began, it was kind of AWS, now you've got other
offerings, but ClickOps is great. You go in, you can,
through a portal, to click what you want, deploy it. And that's great for
experimenting early on, but you have no visibility of what's been deployed centrally. You don't understand who's
doing it in a lot of ways. It's hard to replicate
and redo things like that. So, it's also understand
what's intended to be there, what's not intended to be there. So, ClickOps is great
for the basic beginnings, but if you're really going
to build a program, that's where IaC and GitOps come together,
where you come into play where you can now have a
way to repeatedly deploy, you have structure of what
you expect to be deployed. You can compare that to
what's been deployed. In state files, you've got the visibility. So, those are the major
differences in terms of ClickOps, which has essentially no
real process around it, and then GitOps, which
brings code-like process and GitOps operations, where
you've got pull requests and you've got a code
base you've checked into that describes your infrastructure. So, those are really the differences of one's not a lot of structure. One's highly structured,
a lot of ceremony, and some could say too
much for certain use cases. That is how we view the
difference between ClickOps and GitOps and IaC. >> Yeah, I couldn't agree more.
Rob Strechay
>> I think, again, you guys
had a really interesting announcement back a
couple of weeks ago with that understanding of
what you just went through between ClickOps and
infrastructure-as-code and GitOps. You guys came out with Spacelift in '10. How is that helping bridge that gap between those two worlds?
Dimitri Vlachos
>> What is the impact of that gap? That's really what we're talking about. And we see a couple of main
things, the impact of that gap. One, when developers want to deploy stuff very quickly
for tests, or non-production, or experiments, or even non-critical
applications, the overhead of GitOps and IaC can be quite heavy. You have to do all this process, you have to do a pull request, you have
to write the code, you have to go through these iterations, the feedback loop can be slow. And so, while we love IaC and GitOps, we've built a lot
of our business around helping companies scale that. There's certain use cases
where it's overkill, there's too much ceremony. And so, we introduced Intent,
which is a way of now instead of having to go through
all of this rigor of IaC and writing code and GitOps operations, we basically take some, we've used AI as an approach to do this. So, we allow developers or organizations in
general, whether it's lower- skilled DevOps teams or developers, they can
connect right to an MCP server that we've developed
the Intent MCP server. They're able to now just in plain English or whatever language they're
using, they can describe what they want deployed and basically Intent will
go deploy that for them. And so, you get the speed,
if you will, that ClickOps, not a lot of ceremony, but behind the scenes you, as a DevOps or a platform team, to get
to set a bunch of policy behind the scenes so that
whatever agent, whatever type of LLM your developers or DevOps teams are using, you have control of what it can do. So, it puts guardrails and brings a level of trust around your AI as it's deploying infrastructure. So, that's really what we introduced, and it's a way to allow
developers who want to deploy quickly. We've got a lot of interest
actually from some companies that are looking at,
"Hey, we've got pockets of organizations that are
very advanced when it comes to infrastructure's code. Then, we've got other pockets of teams that are just starting their cloud journey and they don't want to have
to learn all the overhead of IaC, of Terraform. They want to start using
Intent for more of their baseline infrastructure deployment. " So, it's been really
interesting to see the reaction. We really view this as a
really good option to sit side by side with infrastructure's code. We don't believe, and we're not proposing that infrastructure's code and
GitOps should be thrown away. We believe IaC and GitOps is really critical
for critical infrastructure, for known paths, for applications that are absolutely central to you. But for prototyping, for rapid deployment, rapid testing Intent is a perfect fit. So, that's really what we announced and what we've seen with Intent.
Rob Strechay
>> Yeah, I mean I think
that makes total sense for those use cases to be very
complimentary to each other, not in opposition From that
perspective, I think all of this together definitely
brings it home for organizations that want to move fast, but have to have those
guardrails in place. Is that what you're seeing
is this helps them keep those guardrails and really move
fast in that prototyping phase and things of that nature?
Dimitri Vlachos
>> Yeah. If you look at AI and infrastructure, there's
a couple of approaches. One is, "Hey, we're going to
help you write faster IaC code, " but then you have to
put it through your same GitOps operation. And we're also seeing people
have a real trust issue with, "Hey, how am I going to use
AI in my infrastructure? I don't want to let it go hog wild. I don't want to let it in
my critical infrastructure, but I know it's going to come
and I know I need to adapt. " And so, this is where
Intent, you allow AI to be used to deploy parts of your infrastructure, but you have a lot of control over it as a centralized DevOps and platform team and it's almost behind the scenes. The people connecting their LLM to our Intents server don't
even know these policies are there until they try to do
something that's blocked. Then they're told, "Oh, you can't do that. " And the LLM can learn,
"Well, what can I do? " And then deploy what's needed. I think one other thing that's
important to understand, Rob, is we're bypassing code
when it comes to Intent. So, it's not just about
writing code faster, IaC code, but we're basically taking the approach of there's certain cases
where you don't need code. So, what we're doing is going
directly to the actual APIs, the providers, and
skipping the intermediate step of creating code. So, I think those are the
things that people are kind of, ah, interesting. I have the guardrails. I'm able to go faster when I
need, I can start to trust AI with control for certain
parts of my infrastructure. So, that's what we're
hearing from a lot of people.
Rob Strechay
>> Yeah, that makes total sense. And I think, again, both of us were down in Atlanta
for KubeCon + CloudNativeCon. What did you find interesting
in the day zero events, particularly around OpenTofu? I think I didn't make it there,
but I know you guys were big and very involved in there.
Dimitri Vlachos
>> So, with OpenTofu, I think
OpenTofu is starting to reach a very different level of use
and acceptance in the industry. I think it's become more and more, we see a lot of
people converting to it. I think it was the biggest day
in terms even though KubeCon overall the attendance was
down a little, I think a lot because of the travel
restrictions and what have you. But the OpenTofu day was quite packed. The sponsors continue to increase, and the discussion now not only is about what is OpenTofu doing
differently than Terraform, it's also about, well, how
are people deploying it? Where is it being deployed? We've seen Fidelity, Oracle, Cisco, like some major names start to adopt this and offer it as a major side-by- side offering with other types of IC flavors they're supporting
in their infrastructure. So, I think OpenTofu is
just starting to reach more and more of that trust and acceptance as a real viable
IaC option in the market.
Rob Strechay
>> Yeah, I think so too. I mean, there was a lot of buzz about it and I think more buzz than there was in Salt Lake a year ago. And I think it's going
to continue to increase as people become more involved in it, because I think a lot
of it is, like you said, it's not either or, it's an and piece with OpenTofu. But also I was excited I
got to be with you guys at IaCConf Connect on
Monday night down there. It was a great in-person
event, the first in- person event for IaCConf. So, help from the perspective. I was up moderating the panel. You were out in the crowd there. I had great conversations there before and after the moderating the panel. But what was your perspective
on the community actually coming together and what
you've seen as it's grown?
Dimitri Vlachos
>> Yeah, so for those who don't
know, IaCConf is something we started as the sponsor in May. And let me just back up and
say why do we start IaCConf? When you look at this market,
infrastructure-as-code, there's not really a conference, an area where people can come together
in a very open non-vendor heavy environment that is
talking about infrastructure-as- code from best practices,
beginners, advanced use cases, what's on the forefront. And so, we started IaCConf in May. We are the sponsors, but there's very limited space lift promotion at all. It's not a space lift event.
It's really a community event. We had over 3000 people register. It was a virtual event in May,
and we had over 2000 attend. And some of the follow- up feedback was people wanted two things. They wanted spotlight events,
like a particular topic to do a more focused event on. And we did our first one of
those in August on security. And then, the other thing people
really wanted was they want to meet face-to-face. And so, IaCConf Connect came out of that. And so, we do them side-by-side,
in this case, KubeCon. And we had over 200 people register. We had over 75-80 attend,
which was our target, which was great,
especially given everything that was happening with travel
and it was the first night. And the topic of that panel was AI. And so, you moderated, we had some really interesting
different participants. We had Ned In The Cloud. We had a gentleman who was just going through an interview process of... Had left a job and took some time, and then was interviewing
with a bunch of companies to be a DevOps leader and his perspective on what
AI looks like in terms of what people were asking
during his interview. So, it was a really interesting panel. And I think in the room it was really great just to hear people
having a venue to go talk and understand what they're working on, what they're not working on. And what's really interesting
is just the breakdown of the audience We typically
have, it's pretty well split. We ask people during these
events, your skill level at IC, beginner, intermediate, advanced, and it's pretty evenly
split, which is great to see. And we had a number of people volunteer to host their own IaCConf
Connect, so that's great to see that community really
start to continue to grow. And we've got our next IaCConf
spotlight in January on AI, you can see that's a topic
people really want to talk about. So, it felt really good to see
in-person energy in a world that is so virtual today. As you know, rob. You were there. >> Yeah. Yeah, I mean it was great.
Rob Strechay
>> And I know you guys recorded
it as well, the panel. And I would say to people,
go back and watch the panel, because to Dimitri's
point, it was very much non non-vendor specific. It was very much about
how are people dealing with people coming from more
the developer down towards infrastructure and the people
coming from infrastructure up towards the developer. It was a great, I mean, we actually had the cutoff
questions, which was crazy and awesome there. But as you look forward, and we keep going because I think that's
going to be a great thing, but what do you see as
the future for Spacelift as you move forward and what do you hope when
we're together next year again, and actually back in Salt
Lake, that's where it's going to be next year again. Probably the same weather
given how cold it was- >> Oh, that was crazy.
Rob Strechay
>> I thought it was going to
snow there the one day it was
Dimitri Vlachos
>> 29 degrees, it was like... Whatever. But anyway. So, when we're in Salt
Lake, what do you hope to say about Spacelift and
what progress you've made over the course of the year here? >> I think when I look at next
Dimitri Vlachos
>> year, I think a couple of things. One, the whole idea of Intent and IaC, we really see this idea of a two- prong deployment model
where you can use AI- based Intent when it
makes sense, you use IaC and GitOps when it makes sense. So, I would look forward to seeing people adopting
AI in some way in infrastructure to help speed deployment, have DevOps teams focus on the
more critical infrastructure rather than be distracted by use cases where they don't need their expertise and allows the DevOps
teams to move faster. So, that's one piece I would hope comes to fruition, we'll see. The other is with IaCConf,
the continued building of that community, another Connect there at what we will do in Salt Lake City. We're going to do one at Amsterdam. So, I think the continued
building of that community, which is really important
to us as Spacelift. And then, finally, the other component of Intent we didn't talk about
that is it is open source. So, there's an open
source version of Intent, and so we want to see that community build over
the next year as well. So, my wish list for next year
would be those, if you will.
Rob Strechay
>> I love it. If people
want to get more involved with IaCConf, where should
they go and check it out?
Dimitri Vlachos
>> They can just go to iacconf.com
and learn all about it. You can see all of the recordings from prior events, both in-person
and the panel that we had. The panel isn't quite up yet. It will be up in the next
few days, but iacconf. com you can go and see
all our past events, register for current events. We have our CFP open
for our next spotlight, which is in January on AI. So, if you have a good
talk you want to submit, please go ahead and do it.
Rob Strechay
>> Yeah, I can't wait.
I'll be there in Amsterdam and in Salt Lake, so
can't wait to be there. And thanks for coming on,
Dimitri. This has been great. >> Thank you, Rob. Always
great to talk to you.
Rob Strechay
>> And thank you for watching this post-
Dimitri Vlachos
>> KubeCon + CloudNativeCon 2025
episode on infrastructure-as- code on theCUBE, the leader
in tech analysis and news. Stay tuned.
Rob Strechay