Cutting quote times from a day to a few hours with an AI assistant

Too slow to quote

The company handles house moves and storage across two provinces. Quote requests arrive through a form on the website, by email, and sometimes as a photo of a handwritten list of furniture. Two people in the office dealt with them, copying the details into a spreadsheet that worked out the volume, the crew size and the price.

On a busy Monday there could be forty requests waiting. Customers were getting their quote a day or more later, and by then many had already booked with whoever replied first.

Watching how the work was done

We spent two days sitting with the office team. Very little of their time went on the pricing itself, because the spreadsheet handled that. Most of it went on reading messy emails, working out what was being moved and where from, and chasing customers for things they'd left out, like whether there was a lift or how far the van would have to park from the door.

What we built

We built an assistant that reads every new request, whether it's a form, an email or a photo of a list, and pulls out what the spreadsheet needs: both addresses, the date, the rooms and large items, and any access notes. If something important is missing, it drafts a short, friendly email asking for it.

Once it has everything, it fills in the quoting sheet and drafts the reply with the price. A person always reviews the quote and presses send. Nothing reaches a customer without someone in the office checking it first.

Getting it right

The first version was confused by the shorthand the team used every day, and by lists that mixed English and French. We spent the first few weeks going through its mistakes with the office staff and adjusting it until they trusted the drafts. We also made it flag anything it wasn't sure about, rather than guess.

The difference it made

The average time to send a quote dropped from about 26 hours to 3. The office team reckons they've got back around 15 hours a week, which now goes on following up with customers and planning the crews. And because quotes land while people are still comparing their options, 19% more of them are turning into booked jobs.