From a text message to a planned route

In many service businesses, customers book the way they always have: a quick message. “Crush the bins Friday.” “Can we get a swap at the yard this week?” To the customer it is effortless. Behind the scenes, someone has to recognise who is asking, know which site they mean, enter the job, and work out where it fits in a driver’s day.

Proctor is the system we built with a waste-services company to take care of that work. Customers keep messaging as they always have. The booking is recorded in the company’s existing system, and the route adjusts around it. This article walks through how it works.

Understanding everyday language

Regular customers write in shorthand, and a few words carry a lot of meaning. Proctor reads each message and fills in the rest from the company’s own records:

• Who is asking, recognised from the sender.
• Which site, from the customer’s past jobs.
• Which service, from the words they used.
• When. A day with no time means any time that day. No day at all means today.
• The price, which is always the customer’s agreed rate.

Proctor also brings practical knowledge of the trade. A replacement container is the same size as the one already on site, so there is no need to ask. The driver who last visited is assigned again where possible. And if a customer has several sites and has not said which one, Proctor asks, because the right address matters.

Showing its reasoning

Each booking shows the words it was based on. When Proctor decides a message is a request for a particular service, it points to the phrase in the customer’s message that says so, and the system confirms that phrase is really there.

This makes every booking easy to check at a glance, and it is a good habit for any AI system: show the evidence, in the customer’s own words.

Routine requests, and the ones that need a person

Most requests are routine, and Proctor books those directly. Others deserve a second look, and those go to the dispatcher.

Deciding which is which involves both the agent and a set of plain rules. Simple checks run alongside the agent’s own reading of the message. When the two agree and the request is routine, it is booked. When there is any difference, a person reviews it. If a request is waiting and has not been looked at, the owner receives a reminder, so nothing is left sitting.

Requests that arrive in a fixed format, such as a standard form, are read by straightforward code. The simplest reliable method is used for each kind of input.

Careful with every detail

Before a booking is saved, each customer and site reference is checked against the company’s records. Addresses are checked too: they are looked up within the company’s service area, and anything implausibly far away is questioned.

These checks are quiet and quick, and they mean the dispatcher can trust what appears on the board.

A day that adapts

Every new job changes the day a little. Rather than asking whether there is room, Proctor accepts the booking and works out the best route again, which is how a good dispatcher thinks: a regular customer is looked after, and the plan adjusts.

The route is planned from where each truck actually is at that moment. Visits with a fixed time stay exactly where they are. Everything else can move, so a new stop fits in sensibly. If a day becomes very full, the plan allows a little extra time at the end of it.

The dispatcher remains in charge throughout. They can fix a visit’s time, move a job to another driver, and they review and approve each new plan.

Staying dependable

Two small examples show the kind of care that goes into keeping a system like this reliable.

The live map depends on location information from the trucks. Proctor regularly confirms that every truck is visible on the map, which is what the dispatcher actually relies on.

And the system is tested against a collection of realistic messages, so that changes can be made with confidence that everyday requests are still understood correctly.

What changes for the business

For the owner, bookings no longer need retyping, and the day’s route no longer needs rebuilding by hand each time the phone buzzes. For customers, nothing changes at all, which is the point. They send the same short message they always have, and the truck arrives.

It is a good illustration of what a well-designed agent can offer a hands-on business: it takes care of the routine, adapts as the day unfolds, and leaves the people who know the work firmly in charge.