Infrastructure as Code's Hidden Problem with Pavlo Baron

Mar 18, 2026

Pavlo Baron, co-founder and CEO of Platform Engineering Labs and creator of Foray, is a veteran infrastructure and observability engineer. He digs into why Infrastructure as Code still needs babysitting. Topics include Terraform drift and state wrangling, tool sprawl and simpler stacks, type-safe constrained interfaces, use-case-driven guardrails, and how Forme aims to reduce ops toil.

Ask episode

AI Snips

Chapters

Transcript

Episode notes

ANECDOTE

Instana Hotel Hacks Led To Long Term Iac Lessons

Pavlo recounted building Instana from hotel-room prototypes into a company acquired by IBM, which exposed him to large-scale automation challenges.
That experience later motivated him to reimagine IAC tooling with Zach and start Platform Engineering Labs.

INSIGHT

Iac Breaks When It Demands Constant State Babysitting

Infrastructure-as-code tools fail because they force teams to babysit state and chase manual drift instead of supporting real-world use cases.
Pavlo argues Forme focuses on use-case-driven reconciliation and optional enforcement so teams choose when to accept emergency click-ops versus enforcing desired state.

INSIGHT

State Files Centralize Expertise And Block Ownership

Chasing drift and manual state management concentrates complexity on a few experts and prevents team ownership.
Pavlo says state files and wrangling keep ops stuck doing low-level work instead of delivering higher-level platform value.

Get the Snipd Podcast app to discover more snips from this episode

Get the app

Terraform drift, state wrangling, and a growing “tools for tools” stack are still daily work for many platform teams - despite a decade of DevOps talk and cloud maturity. Why does ops automation so often feel like it needs babysitting?

Pavlo Baron breaks down where Infrastructure as Code tends to break down in real organizations: manual drift management, low-level state complexity, and a lack of practical abstractions that let developers self-serve without inheriting the entire ops burden.

The conversation digs into what a more use-case-driven approach could look like - where teams can choose when to enforce desired state, when to accept emergency changes, and how to build “guardrails” that reduce mistakes without slowing delivery.

Pavlo also explains why type safety and constrained interfaces matter (especially as AI starts generating more code and infrastructure changes), and why the future of platform engineering depends less on slogans and more on systems that reduce toil.

Guest: Pavlo Baron, Co-Founder and CEO of Platform Engineering Labs

Pavlo Baron is Co-Founder and CEO of Platform Engineering Labs, who are crafting tools to remove the toil from the operations work, with a current focus on infrastructure. He is a veteran in the space, having served in all kinds of roles throughout his career that spans more than 35 years. Previously, he was co-founder, CTO, and major inventor at an observability startup, Instana, that was acquired by IBM in 2020. Pavlo is a frequent conference speaker and author of several books.

Pavlo Baron, X

https://pavlobaron.medium.com/

https://github.com/platform-engineering-labs

https://www.linkedin.com/company/platform-engineering-labs

https://x.com/plateng_labs

https://bsky.app/profile/platform.engineering

https://mastodon.social/@plateng_labs

https://www.youtube.com/@plateng-labs

Links to interesting things from this episode: