What if, in five years, the system builds itself — a product manager describes what they want, an architect agent inspects the existing codebase, and the feature gets written without a traditional dev team in the loop? In this episode, Erik Parienty, Director of DevOps Engineering at Riskified, lays out that future and what it means for the people running infrastructure today. Erik has spent 7 years scaling Riskified's DevOps from a 6-person team through the company's IPO, and he's been in the field since long before the word "DevOps" existed — which gives him a rare long-lens view on what he calls the discipline's "second revolution."

Erik Parienty on LinkedIn - https://www.linkedin.com/in/erik-parienty

Miki Manor on LinkedIn - https://www.linkedin.com/in/mikimanor/

Ofir Stein on LinkedIn - https://www.linkedin.com/in/ofir-stein/

DevOps Leaders - https://www.linkedin.com/company/devops-leaders-il/posts/?feedView=all

Key takeaways:

• The "second revolution": the first wave standardized CI/CD and Kubernetes — the next automates the entire SDLC and PDLC, from the product idea forward

• The future org chart: a product manager talks to an architect agent that reads your system and decides whether to customize what exists or build something new

• Stop blaming the model — Erik's reframe: "if you told it and it didn't do it, you don't understand it." The agent is a tool; agent literacy is the new Google search

• The rise of the one-person company — why you no longer need to raise a round to build, and why (like the Google panic) revolutions create more work, not less

• The FinOps cautionary tale: an agent insists your $2M Kubernetes cluster costs $50, because it pulled context from a random server tagged "kubernetes"

• Where DevOps still wins: human adaptability. The communication skills we use with people are exactly what make someone good at operating agents

• The coming open-source agent ecosystem — git clone a community-maintained agent, run it locally, and it just does the work, the way we run Karpenter and cluster-autoscaler today

• The measurement problem nobody's solved: from counting lines of code at Amdocs to counting tokens — and what that says about productivity

• What AI still can't touch — infrastructure. Root-cause analysis across distributed systems is the real frontier, well beyond today's self-healing watchdogs

• We're recording live from the PlatforMa conference in Tel Aviv, featuring fascinating discussions on AI and more.

Join our exclusive DevOps Leaders community with over 300 members, where we share knowledge and improve together.

What's your biggest challenge in the DevOps field? Drop it in the comments! Subscribe for weekly insights and updates on the latest in DevOps!

#DevOps #Podcast #Technology #AI #Riskified