The problem
Enquiries arrived with multi-sheet bills of quantities covering several brands. An estimator had to read each line, find the matching item in the catalogue, apply the right price and discount rules, and build the quote by hand.
At one to three days per quote, the delay was not the estimator's speed. It was the volume of lookups and the queue of enquiries waiting their turn, and slow quotes lose work to faster competitors.
How the agent works
The design splits judgement from arithmetic, because a model is good at reading messy input and a rules engine is good at getting the same price every time.
- The agent reads the enquiry and every sheet of the bill of quantities, whatever the layout.
- It matches each line to an item in the live ERP catalogue across the brands in the schedule.
- A deterministic rules engine applies prices and pricing rules, so every figure on the quote can be traced to a catalogue entry and a rule.
- The draft quote goes to a person to review and send.
The pricing logic sits behind a suite of 87 tests, so a change to a rule or a model cannot quietly change a price.
What changed
Quote turnaround went from one to three days to under three hours, about 90 percent faster, priced against the live catalogue rather than a copy that drifts out of date.
What a project like this needs from you
Access to the systems the work already lives in, usually a mailbox and the system of record, with a login the agent can use safely. The rules written down, or a person who knows them and has an afternoon to explain them.
A set of real past cases to test against. Replaying last month through a new version before it goes live is what lets a team trust the change. A named owner for the decisions the agent will not make, and a period where it runs alongside the team in shadow mode before it acts on anything.