Posts

Agents are non-deterministic. Security assumed determinism. Sandboxing and isolation aren't just for AI; they're the only path when we can't policy our way out.

In the age of AI agents, we're not just shifting left for humans. We're shifting left so agents can run and trust CI themselves. That means local CI that's fast enough to iterate on.

Why the shift to 'post-agentic' workflows is inevitable, and how the economics of cheap intelligence is fundamentally changing how we build software.

Reflecting on my initial assumptions about AI agent security, filesystem isolation, and why simplicity often beats over-engineering.

How I moved beyond traditional patching to build a 0-CVE, immutable operating system for VM workloads using RHEL image mode.

Docker Inc's journey from containerization pioneer to AI-focused platform has been anything but straightforward. Here's what their strategic shifts tell us about the company's search for identity.