AI-DevOps consulting
AI agents that don't
blow up your stack.
I help engineering teams adopt AI coding agents safely — with the cost controls, IaC safety gates, and CI discipline of someone who ran production systems for 25 years.
the problem
Three things going wrong right now
- Your AI coding bill is leaking. No spend visibility, no model right-sizing, no context controls. Token waste is invisible until the invoice arrives.
-
Your agents have no guardrails.
Claude Code can propose an
applyon your Terraform state. Nothing stops it. One yes in the wrong window is a real incident. - Your team adopted the tool, not the workflow. Vibe coding scales individual speed; it doesn't scale team quality. Reviews, gates, and discipline don't happen by default.
proof
See the system in practice
A detailed teardown of the harness I run: cost-governance controls, IaC safety gates, six-phase PR-review skill, and what didn't survive. Real config, real numbers — no marketing.
Read the teardown →rates · USD
How we work together
Launch-rate engagements available for the first client cohort, in exchange for a case study.
free resource
12 things your agentic config needs before it touches IaC
If you can't tick every box, the agent is going to break something a code reviewer can't catch. One page. Free.
book a call
30-minute intro call
Monday, Wednesday, and Friday mornings, Ottawa time. No pitch — a conversation about what your team is doing with AI agents and where the friction is.