Blog

Agents and models are becoming commodities. The enterprise advantage now lies in the operating model, governance, context, and accountability around them.

Packaging a Google Cloud delivery team as an installable agent plug-in — the reusable practice knowledge that otherwise gets rebuilt on every engagement.

The demo-to-production gap: ownership, operations, governance, run-cost, and the 2am pager problem no pilot ever tests for.

How human-agent team composition changes when every SME owns a squad of agents instead of a headcount line.

Two years of running a prompt-management extension, rebuilt on Chrome's built-in on-device AI — what privacy-first actually costs and what it buys.

Co-authored. Maturity from discovery to run, the Day 2 problem, the AI Centre of Excellence, and context engineering as the durable advantage.

Agentic AI has changed what small teams can produce. Now organizations need the harness, enablement, and operating model to keep up.

One-off agents feel reasonable until identity, access, governance, observability, and ownership turn into platform sprawl.

Building a prompt-management extension end to end — why prompt reuse breaks down in practice, and what a managed library fixes.

Building a C4-model architecture diagram extension end-to-end with Gemini 3 and Google's Antigravity IDE — a weekend project that shipped.

Building an enterprise-grade, privacy-first Slack summarisation agent on Google Cloud — the beta release write-up.

The relationship between developer experience, SRE, and a Cloud Centre of Excellence — a strategy for platform rollout in the enterprise.

The write-up behind the first published extension — breaking meetings and deep work into focused intervals.

Navigating multi-cloud strategy — flexibility, operational overhead, and security trade-offs, from someone who has run more than one cloud at once.

A practical guide to fortifying container platforms — tools, projects, and best practices to mitigate risk and hold compliance.