Publishing consistently is a production problem, not a writing problem. The bottleneck is never the first draft. It is the artwork, the social variants, the metadata, the upload, and the discipline to do all of it the same way every time. We built the pipeline that handles the parts around the writing.
A repeatable pipeline that takes a tax or accounting topic from research to a published article with its own artwork, social assets and search metadata, in the same shape every time.
Every piece lives in a self-contained folder with the same files in the same places. Nothing lives only in someone’s head.
Explaining a tax threshold needs a diagram, not a photograph of a handshake.
AI drafts. A person decides what gets published, and takes responsibility for it.
The last mile, getting the piece and its assets into the site correctly, is automated.
Most of the value in a system like this comes from the systems it talks to, and most of the risk does too. Each connection is scoped to the access it needs for the job.
The path a piece of work takes through the system, start to finish.
A topic is picked against what the audience is actually searching for, and a brief is written covering the angle, the audience and the questions the piece has to answer.
AI assists with research and the first pass. A person edits it into something worth reading, checks every figure and date, and takes responsibility for what it says.
The image brief becomes a set of custom vector graphics (a thumbnail, explanatory infographics and a social asset) rendered to web formats.
The edited copy, the artwork and the metadata are assembled with the upload manifest into one folder, ready to publish in a single pass.
Tooling uploads the piece and every asset in a single pass, with the metadata applied from the manifest.
The problems that decided how this was built. They are the reason it works the way it does rather than the obvious way.
AI-drafted content is fast and, left alone, uniformly bland. In tax and accounting it is also occasionally and confidently wrong, which on a regulated firm’s site is a real problem rather than an embarrassing one.
AI is used where it is genuinely good (research, structure, first drafts) and a person does the editing that makes it worth reading. Every figure, threshold and date is checked against the primary source, and the piece is edited into a real voice before it ships.
A defined pipeline with editorial checkpoints, and a named person accountable for every claim in every article.
We don’t publish screenshots, logos or client names for this work. Much of it runs inside a business and handles sensitive data, so we describe what was built rather than who it was built for. We’re happy to walk through the detail on a call.
Every build on this page started as a workflow that off-the-shelf software nearly handled. Tell us where the gap is and we’ll scope what it takes to close it.