Bidwin is an AI-powered drafting tool. It generates structured, evidence-cited first draft bid responses based on your own tender documents, evidence files and writing style, not generic AI output.

Generic AI tools like Claude and ChatGPT can draft text, but they can compromise your data and intellectual property, and they lack your evidence files, bid history, and scoring-criteria awareness. Bidwin-AI solves this by building a custom drafting system inside your Claude account, grounded in your own documents so every draft is built on real evidence, not hallucinated content.

Bidwin Watch is our tender opportunity finder. It scans procurement portals for live bid and tender opportunities, scores each one against your own evidence, and tells you what is worth bidding. For a limited time, founding members receive it free (normally £199 a year)

No. Your £349 is a one-off payment and the licence is yours permanently. Nothing auto-renews and we take no further payment without you asking for it. If you never renew, you keep the version of the plugin you have and carry on using it. What a renewal buys is another twelve months of updates and support, which is worth having because Bidwin runs on Claude and Claude keeps moving. The one genuine subscription in all this is your own Claude Pro account, which Anthropic bills directly.

Two reasons.
The first is where your files live. Bidwin installs into your own Claude account and works from folders on your own computer. Your tender pack, your evidence library and your drafts stay with you. We never hold them and we cannot read them. Most AI writing tools ask you to upload your material to their servers first, which isn’t something that every organisation is comfortable doing.
The second is that Claude lets us install a method rather than hand you prompts. Mark’s drafting discipline, the tender vetting, the evidence rules and the pre-submission review come as a plugin that runs the same way every time.

Bidwin never sees your bid content. We do not hold your library and we cannot read your drafts. That moves the question along to Anthropic, where on an individual plan it comes down to a single setting, and Help Improve Claude is not off until you switch it off. Turn it off under Settings then Privacy, and new sessions are not used to train future models. Do that before your first tender, not after

It can, and every one of the causes is avoidable. Costs run away when someone keeps a single conversation open for a fortnight, attaches their whole evidence library to every question, and runs the heaviest model on routine edits. Claude’s usage is visible rather than mysterious: a rolling five-hour session, a separate weekly ceiling, and both shown under Settings then Usage. Bidwin is built to work with that rhythm. It steers you to the lighter model for routine sections and the stronger one only where it earns its place, offers a quicker tender-vet mode when you don’t need the full analysis, and works from one scoped tender folder rather than your whole drive.