Small team.
Senior engineers.
No fluff.

I started Webbyfox after a decade of building digital products inside agencies, startups and in-house teams. The pattern I kept hitting was the same: a project would be "done", handed over to the client, and six months later someone would be on the phone asking why the plugin that runs half the site just had a critical CVE, or why a simple content change needed a developer, or why the migration data that was supposed to be in a field was actually buried in a WordPress shortcode.

The common thread wasn't bad intentions. It was structural choices made early that locked teams into ongoing complexity: CMSes that couldn't enforce schema, page builders that let every editor invent a new layout, bespoke plugins that nobody understood. The maintenance burden compounds until it's a significant line in the engineering budget every quarter.

Wagtail changed how I thought about CMS work. Here was a platform built by engineers who took content modelling seriously — typed StreamField blocks, version history, page-tree structure, and a Django foundation that put the database one import away from your data and AI tooling. After the first two Wagtail projects, I stopped taking WordPress work.

The AI practice grew organically from the same frustration: off-the-shelf LLM wrappers that require constant prompt-fiddling and fall over when the underlying model changes. I started building retrieval systems — pgvector, LangGraph, structured RAG pipelines — that treat the AI layer as an engineering concern, not a product feature bolted on late. The data is typed, the retrieval is testable, the cost is predictable.

Webbyfox is deliberately small and deliberately senior. We take on 3–5 projects at a time, work alongside your team rather than in a black box, and write code that your engineers can reason about after we've left.