vertical ai

turning material specs into quotes.

a wa piping supplier priced every job off a specification that arrived as a pdf or a spreadsheet, and somebody had to read it before anything could be quoted.

up a level

/case-studies

see also

/ai-workflows/industry/property-services/method
tower cranes over a half-built concrete frame on a ridge, the coast beyond
a site outside perth. everything on it was priced off a specification somebody had to read first.

the situation.

customers send a material specification as a pdf or a spreadsheet, and no two of them are laid out the same way. staff read each one, keyed the line items into their system, then chased supplier quotes by hand before anything could go back to the customer.

it was slow at the best of times, and at the worst of times it was the reason a quote went out a day late.

the client

a piping supplier in western australia. the name is withheld until they have agreed to be named.

what we built.

an agent that reads the specification whatever shape it arrives in, pulls out the line items, matches them against the catalogue, and drafts the quote in the system the team already uses.

it does not send anything. a person reviews the draft and makes the call on price, which is the part that was never the bottleneck.

human in the loop

the agent drafts. a person prices and sends. that line was set before we wrote any code.

what changed.

the reading and the keying went. what is left for the team is the judgement: whether the match is right, and what the job is worth.

talk to us.

bring us the problem. we will tell you what we would build and how long it takes.

book a call