Quotedesk — a quoting marketplace for sales teams

The UK sales organisation of an international group quoted the way most sales teams do: spreadsheets, supplier price lists of uncertain age, and approvals by e-mail. Every quote took longer than it should have, and nobody could say with confidence that the prices in it were still right.

Quotedesk replaces that. Sales reps search the supplier catalogs, put the quote together, get it approved when it crosses their business unit's limit, and send the customer a finished PDF or Excel file. The catalogs come from the customer's procurement platform, which we built as well, so prices follow the supplier's updates and nobody maintains them twice.

What it does for the sales team

A quote that used to take an afternoon of copying now takes minutes. The approval rules the company already had are now part of the tool: below the limit a rep sends straight away, above it a manager approves first, and nobody approves their own quote. Every quote has a full history, and a changed quote goes out as a numbered revision, so the customer always knows which version counts. The app speaks English, German, French and Spanish and signs people in with the company's Microsoft accounts.

Next: AI that thinks along

The next stage brings AI into the quote. The rep describes what the customer needs, and the AI finds the matching items across the supplier catalogs. It learns from past quotes which items belong together and what similar customers received, and drafts the new quote from there. The rep still decides. The AI takes over the searching.

Two weeks to production

The project started on 22 September 2026. Two weeks later Quotedesk ran in production, with the full scope a company needs before it hands a tool to its sales force: user administration, approvals with e-mail notifications, revisions, and quote documents in the customer's own format. It was built to the standards an enterprise customer expects, with security reviewed against the OWASP Application Security Verification Standard at Level 2.

How we work

Speed came from clear decisions, not shortcuts. Every open question went to the customer with the options and their trade-offs, and most were settled within a day. AI coding agents do much of the implementation under written rules, while everything that touches money, access or security waits for a human decision. That combination lets a small, senior team deliver at the pace of a much larger one.

If your company needs something like this, or someone to set up how such a product gets built, write to us below.