Notes on cloud architecture and AI systems, as I build them.

This section is under construction. The plan is short, specific posts drawn from real architecture decisions — not summaries of documentation you could read yourself.

cloud architecture

Where OCI and Azure IAM models actually diverge

A practical comparison for anyone moving between the two, past what the docs cover.

ai systems

Building a multi-agent pipeline that doesn't fall over

Lessons from a real agentic project — orchestration, failure modes, and where a single LLM call would have been fine.

enterprise patterns

Strangler fig, in practice, on a system nobody wanted to touch

What the pattern looks like once it meets a legacy system with real stakeholders and no downtime window.

Posts aren't live yet. If you'd like a note when the first one is, reach out directly.

Get in touch