Whoever prepares a custom quote today searches their memory for a similar job, opens it, adapts it and redoes the maths. The assistant handles the first part: it reads the request, finds the closest past jobs and proposes an estimate that the person corrects and signs.
When you work to order, every request is different: a forty-page specification, a dimensioned drawing, an email with three lines of requirements. Preparing the quote takes days, and only two or three people in the company know how to do it well.
Meanwhile the customer waits, and has often asked others for the same quote. And people estimating in a hurry tend to be wrong in the same direction, too low: the margin is discovered once the job is closed.
Specifications, drawings, requirements, emails. It pulls out requirements, quantities, materials and constraints, plus the unclear points to clarify with the customer.
It searches closed jobs for the closest ones, with their actual times and costs rather than just the quoted ones. That is where the value lies: in a history that nobody can consult today.
Line items, quantities, hours and prices from up-to-date price lists, with the past jobs each number comes from next to it. Whoever prepares the quote corrects, adds and decides.
The quote in the company's format, with the standard terms and notes on open points, ready to review and send.
These are the numbers we set in discovery, before writing any code. The starting value is measured on your data, and the prototype is judged against it.
From the request to sending it to the customer. It decides how many requests can be followed up.
Once the job is closed, how far the quote was from the real cost. It tells you whether estimates are actually improving.
How many quote requests get an answer, instead of sitting in a queue until the customer goes elsewhere.
If there is no job history, or it has no actual costs, the assistant has nothing to rely on and its estimates are worth as much as a guess. And if the products are catalogue items with fixed prices, you need a configurator, not AI.
The info@ inbox receives everything and someone has to sort it by hand. AI reads every email, gives it a category and forwards it to the right department. It replies to nobody: it routes.
Delivery notes, supplier confirmations, contracts, certificates. AI recognises every document, extracts the data for the ERP and flags what a person needs to look at.
A free 30-minute call, with no commercial follow-up unless you want it.