Case study · Freight forwarding

AI quoting for a US freight forwarder

Quotes were assembled by hand from rate sheets and email threads. We built a system that drafts them, so the team reviews instead of rebuilding.

Freight
Industry
US
Region
AI
Service

The challenge

Every inquiry started the same way: read the email, find the lane, open the rate sheets, add surcharges, build the quote. The work was repetitive but easy to get wrong, and response time depended on who was free. The client needed faster quotes without handing pricing decisions to a black box.

Approach

What we did

01

Map the workflow

With the sales team, including the exceptions they handle by instinct.

02

Structure the rates

So lanes, carriers, and surcharges can be matched reliably.

03

Build the AI layer

Reads inquiries and produces a quote draft with its rate sources shown.

04

Keep people in control

Every draft is reviewed and edited before it goes to a customer.

Technical

Architecture

01

Inquiry parsing

An LLM extracts lane, cargo, and service details from free-form requests.

02

Rate database

Structured rates and surcharges, replacing scattered spreadsheets.

03

Traceable drafts

Each quote line shows which rate it came from.

04

Review interface

Sales edits and approves drafts in one place.

Related

A system to build,
or one that needs taking over?

Book a consultation →