Writing

Notes on software & AI.

Practical write-ups from the work — architecture, integrations, and where AI actually earns its place. Sample posts for now; real ones to follow.

Token lifecycles, per-client WABAs, and keeping a single webhook endpoint sane across many businesses.

What it actually takes to put a useful retrieval layer behind an LLM without standing up a whole platform.

Auto-configuring devices on boot — config templating, cloud APIs, and the gotchas that bite in production.

Why one model rarely fits a whole product, and a simple way to route each task to the model that suits it.

The licensing maze, the trade-offs between stacks, and why the audio pipeline is the part that bites.

Cutting through the hype — the handful of spots where AI genuinely moves the needle, and where it doesn't.

Want these in your inbox?

We write when we have something worth saying. No noise — just notes from real builds.