Guest: Dimitri Vlachos (LinkedIn)
Company: Spacelift
Show: 2026 Predictions
Topic: Infrastructure as Code

The velocity gap between development and infrastructure teams is widening. As developers accelerate code creation with AI assistance, platform and DevOps teams are struggling to maintain pace—and traditional approaches like GitOps and Infrastructure as Code (IaC) alone won’t be enough to bridge the gap. Dimitri Vlachos, VP Marketing at Spacelift, shares critical predictions for 2026 and actionable strategies for platform teams navigating this transformation.

The Acceleration Crisis Facing Platform Teams

Spacelift helps DevOps and platform teams deploy, manage, and automate infrastructure management. But 2026 brings a fundamental challenge: the speed at which developers are creating code with AI assistance is skyrocketing, and infrastructure teams are at risk of becoming bottlenecks.

“It’s getting harder and harder for platform teams to keep pace with developers,” Vlachos explains. “The speed at which they’re developing code with AI assistance is just skyrocketing.”

This acceleration creates tension. While development teams move faster, platform teams remain responsible when systems fail. The challenge isn’t just about speed—it’s about maintaining governance, security, and reliability while enabling unprecedented velocity.

Four Critical Predictions for Infrastructure Teams

Vlachos identifies four key trends that will define platform engineering in 2026:

First, infrastructure teams must adopt AI themselves. “We believe that IaC alone, like traditional GitOps pipelines, is not enough to keep pace,” he notes. “It’s an important piece, but infrastructure teams need to start looking at how they can adopt AI to keep pace.”

Second, governance and security will become more critical, not less. “With demands for more velocity, governance becomes even more important,” Vlachos emphasizes. “How do you go faster but ensure you’re not putting resilience, security, and reliability at risk?”

Third, natural language AI interfaces will become integral to daily workflows. Just as AI has transformed development processes, Vlachos predicts that AI chat and natural language interfaces will become a meaningful part of how platform and DevOps teams work.

Fourth, companies are looking for ways to accelerate beyond traditional approaches. “We’re seeing more and more companies ask, ‘Is there a way for me to move faster without even having to learn HCL or IaC? Can I find another way to automate?’” Vlachos explains.

Balancing Experimentation with Control

One of the biggest challenges platform teams face is perceived as roadblocks rather than enablers. The solution, according to Vlachos, lies in creating flexible deployment models that allow speed where appropriate while maintaining rigor for production systems.

“How do you experiment more without putting yourself at risk?” he asks. “How do you allow people to prototype faster while still maintaining control in production?”

This approach requires platform teams to rethink their deployment strategies—flexing and moving fast where experimentation is needed, while bringing rigorous governance to production deployments.

The Governance Challenge for AI Adoption

As platform teams adopt AI tools, governance frameworks must evolve beyond traditional deployment policies. Vlachos stresses the importance of establishing clear guidelines for AI usage in infrastructure management.

“How do you use governance not just for how you’re going to deploy, but also to bring that same structure to how you use AI in your infrastructure?” he asks. “Whether that’s prompt guidelines or clear rules about where you’re going to use AI and where you won’t, each company will follow a different path.”

AI as a Force Multiplier

The opportunity for platform teams extends beyond deployment acceleration. AI can help teams better understand existing infrastructure—a critical capability given the complexity of long deployment histories.

“A lot of times, when you deploy, there’s a long history of what’s already been deployed, and really understanding what’s there—and where new things are going to fit in—can be quite hard,” Vlachos notes. “How can you start to leverage tools like AI and its power to understand what’s deployed and what will fit best into your environment?”

He describes this as having an assistant for designing, understanding, and architecting infrastructure—both for read-only access to comprehend existing systems and for actual deployment capabilities.

Actionable Advice for Enterprise Leaders

Vlachos offers concrete guidance for platform leaders preparing for 2026:

First, define your guardrails. “What are the guardrails you want to put in place?” he asks. “Where do you want to actually increase your velocity? Where do you want to make trade-offs between speed and control?”

Second, think intentionally about where AI fits. Don’t just dive in—establish a clear framework for where you’ll use AI for understanding and designing versus actual deploying.

Third, remember accountability. “When something goes down in the middle of the night, it’s not the dev team that gets called,” Vlachos points out. “It’s the platform team—the DevOps teams, the SRE teams—that are responsible.”

Spacelift’s 2026 Focus

Spacelift is building on its Intent product, released in late 2025, which enables natural language AI-based infrastructure deployment without requiring HCL or code creation. The company is developing an AI assistant that sits alongside teams across every aspect of infrastructure management—from visibility and policy building to deployment and design.

“We want to build a structure and approach to modern infrastructure automation that gives you the rigor of IaC and GitOps, while complementing it with AI and supplemental intelligence—so you can move fast when you need to,” Vlachos explains.

The goal isn’t to replace traditional infrastructure as code approaches but to create a modern framework that combines the rigor of GitOps with the speed and intelligence of AI assistance.

As 2026 unfolds, platform teams that successfully balance velocity with governance, experimentation with control, and automation with intelligence will transform from perceived bottlenecks into essential enablers of organizational acceleration.