As AI coding assistants reshape the pace of software development, platform engineering teams are under mounting pressure to keep up — and the infrastructure tooling underneath them is struggling to match that velocity. Ahead of KubeCon EU 2026, Spacelift is stepping into that gap with a significant announcement: Spacelift Intelligence, an AI-native operating model built to help infrastructure and platform teams manage the full lifecycle of both traditional Infrastructure as Code and AI-provisioned resources. The launch brings together Spacelift Intent, now generally available, and a new conversational AI assistant spanning the entire infrastructure lifecycle — from provisioning to drift detection to diagnosing failed runs.
VMblog sat down with Dimitri Vlachos, Chief Marketing Officer at Spacelift, ahead of the show to get into the details behind the announcement, the architectural philosophy driving the platform, and how early design partners like Cityblock Health and Vega are already putting Spacelift Intent to work. Vlachos makes a compelling case that AI and Infrastructure as Code are not competing forces but complementary ones — and that the organizations succeeding right now are the ones treating them that way. Attendees can see it all live at Booth #796.
++
VMblog: Are you launching any new products, unveiling significant features, or announcing partnerships at KubeCon EU 2026? What should attendees watch for?
Dimitri Vlachos: We’re announcing Spacelift Intelligence, a new AI-native operating model designed to help infrastructure and platform teams keep pace with the speed at which AI-powered software development now moves. Spacelift Intelligence brings together two capabilities: Spacelift Intent, which is now generally available, and a new AI assistant that spans the entire infrastructure lifecycle.
Spacelift Intent gives developers and platform teams a natural language path to provision infrastructure without needing to understand the underlying cloud constructs. Policy guardrails enforce what’s allowed, audit trails capture everything, and platform teams retain full visibility and control. The AI assistant extends this further, allowing teams to query their infrastructure state, diagnose failed runs, detect drift, and create policies, all through a conversational interface.
KubeCon EU is our first event opportunity to show this live, and we’re looking forward to demonstrating both capabilities at Booth #796.
VMblog: What’s your elevator pitch for KubeCon EU 2026? If attendees only remember one thing about your company after visiting your booth, what should it be?
Vlachos: Developers are now spending roughly a quarter of their workday working alongside AI coding assistants, according to Google Cloud’s 2025 DORA report. The volume and velocity of infrastructure requests that creates is something traditional IaC pipelines were never designed to absorb. Platform teams are being asked to respond at the speed of experimentation while still maintaining the security, compliance, and operational consistency their organizations depend on.
Spacelift is built for this reality. We help platform teams manage the full lifecycle of both traditional IaC and AI-provisioned infrastructure, with an intelligence layer that gives them visibility and control across both. If attendees take one thing away from visiting our booth, it should be that AI and Infrastructure as Code are not in competition. The organizations succeeding right now are the ones treating them as complementary, and Spacelift is the platform that makes that possible.
VMblog: Let’s get technical — what specific architectural challenges or operational bottlenecks does your solution address for cloud-native teams in 2026?
Vlachos: The core challenge is a widening gap between how fast developers can ship and how fast infrastructure teams can respond. AI coding tools have dramatically accelerated the software development cycle, but the infrastructure layer has not kept pace. Platform teams are still operating in a model built for a slower world, and the pressure that creates is showing up as deployment delays, policy workarounds, and untracked infrastructure sprawl.
One instinct has been to apply AI to Infrastructure as Code itself, essentially using it to generate Terraform faster. This approach compounds the problem rather than solving it. It increases the volume of code flowing into pipelines that are already overwhelmed, without addressing the underlying bottleneck.
The architectural answer we’ve built toward is a dual-deployment model. Natural language provisioning through Spacelift Intent handles rapid prototyping and experimentation, with governance baked in from the start rather than applied after the fact. Traditional IaC and GitOps workflows remain the system of record for production. Atop both sits the Spacelift Intelligence layer, which gives teams unified visibility and operational insight across the entire environment. The goal is to eliminate the tradeoff between speed and control, because in practice, platform teams shouldn’t have to choose between them.
VMblog: Platform engineering continues reshaping how organizations approach cloud-native infrastructure. What’s your perspective on this shift, and how does your technology enable effective platform teams?
Vlachos: Platform engineering has moved from a niche practice to a recognized discipline because organizations have learned that giving developers self-service access to infrastructure, without the right guardrails and visibility, creates as many problems as it solves. The teams we work with have lived that lesson. They’ve built out self-service capabilities only to find themselves managing a sprawling surface area of infrastructure that is difficult to track, govern, or safely promote to production.
What’s changed in the past year is that AI has raised the stakes considerably. The volume of infrastructure being requested is growing faster than platform teams can absorb it using traditional processes alone.
The platform teams navigating this well are the ones treating AI as a way to amplify their expertise rather than route around it. Spacelift Intent lets developers move quickly through a natural language interface, while the policies and constraints the platform team has defined continue to govern what gets deployed. The platform team’s knowledge is encoded once and applied consistently, rather than requiring their involvement in every individual request. That’s what allows platform teams to scale their impact without scaling their headcount proportionally.
VMblog: Security, compliance, and governance remain top concerns. What’s your approach to helping organizations build secure, compliant cloud-native systems without sacrificing velocity?
Vlachos: Our starting point is that governance and velocity are only in tension when governance is applied as an afterthought. When compliance requirements are encoded into the deployment model itself, speed and control can coexist.
In practice, that means platform teams define their policies, approved patterns, and constraints once within Spacelift. From that point forward, whether a developer is provisioning infrastructure through natural language via Spacelift Intent or through a traditional IaC pipeline, those guardrails are enforced automatically. Developers gain faster access to the infrastructure they need, and platform teams maintain confidence that what’s being deployed meets their organization’s requirements.
The Spacelift Intelligence Infra Assistant extends this further by giving teams continuous visibility into their environment. Drift detection, diagnostic analysis of failed runs, and natural language queries about infrastructure state all reduce the operational burden that typically forces teams to choose between moving fast and staying compliant. We want compliance to be the path of least resistance, not an obstacle that developers are incentivized to work around.
VMblog: AI and machine learning workloads are becoming first-class citizens in Kubernetes environments. How does your solution address the unique infrastructure and operational demands of AI/ML pipelines?
Vlachos: AI and ML workloads put particular pressure on infrastructure teams because they tend to be resource-intensive, rapidly iterative, and highly experimental by nature. Data scientists and ML engineers need environments they can spin up and tear down quickly, often without deep cloud expertise. At the same time, the infrastructure supporting those workloads still needs to meet the same security and compliance requirements as anything else in the organization.
That’s a version of the broader challenge Spacelift Intelligence is designed to address. Spacelift Intent is particularly well suited to the experimentation phase of AI and ML development, giving teams a fast, governed path to provision the infrastructure they need without requiring platform team involvement at every step. The policy guardrails ensure that even rapid provisioning stays within approved boundaries, and the audit trail provides the visibility that compliance and security teams require.
As AI workloads mature from experimentation into production, the ability to promote Intent-created resources into traditional IaC and GitOps workflows means organizations don’t have to rebuild their infrastructure from scratch when they’re ready to scale.
VMblog: Can you share a compelling customer story or use case that demonstrates tangible outcomes?
Vlachos: Two of our early access design partners offer perspectives that reflect the range of ways teams are putting Spacelift Intent to work.
At Cityblock Health, the challenge was not a lack of self-service infrastructure access but what happened after developers used it. Engineers could spin up development environments quickly, but the work beyond initial setup remained largely manual, creating a growing surface area of infrastructure that was difficult for the platform team to track or safely promote to production. Intent changed that interaction model by letting developers describe what they needed in natural language while platform-defined policies handled governance behind the scenes.
At Vega, Platform Lead Joe Hutchinson approached Intent with a specific goal: enabling experimentation without expanding operational risk. The team ran Intent in an isolated environment with role-based access controls, creating a safe space for engineers to explore ideas without touching production systems. Joe’s observation was that clarity of purpose drove adoption: “It works well when people understand what it’s for. Once you frame it as a place to explore ideas quickly, without touching production systems, the workflow makes sense and adoption becomes much easier.”
Both stories point to the same underlying conclusion. AI-assisted infrastructure works best when it’s paired with strong guardrails and a clear path back to governed, versioned GitOps pipelines.
VMblog: Where exactly can attendees find you on the show floor, and what’s the one must-see demo or experience you’ve built that showcases your technology’s real-world impact?
Vlachos: Attendees can find us at Booth #796. We’d encourage anyone who works on or with platform teams to stop by, because the demo we’re running speaks directly to the day-to-day reality those teams are living.
The demonstration walks through Spacelift Intelligence in action: a developer describing an infrastructure need in plain language, the system applying platform-defined policy guardrails in real time, and the platform team maintaining full visibility and control throughout. We also show the AI assistant at work across the broader infrastructure lifecycle, from querying environment state to diagnosing a failed run to detecting drift, all without requiring anyone to dig through logs or code.
We’ll have the team there to walk through live scenarios and discuss how this applies to your specific environment. Come pay us a visit at Booth #796!
##
