AI agent for freight order intake

Petaron logo mark

PETARON

Order entry, done by an agent.

The problem

Orders still arrive

as emails.

Someone opens each one, reads the attachments and retypes it into the TMS by hand. Hours of copy and paste, typos that show up later, orders that wait after hours.

A mail inbox full of booking requests from customers
The Petaron order queue: orders completed by the agent, one processing, one marked as no order

What Petaron does

From inbox

to a completed order.

An AI agent reads every email and its attachments, sorts orders from everything else and drafts the order before anyone opens it.

How it works

Four steps.

One person in control.

01

Read

Every email and every attachment: PDFs, scans, spreadsheets, forwarded mail.

02

Understand

A new order, an update, a cancellation, or not an order at all.

03

Match

Customers and sites to your TMS codes, completed from your order history.

04

Approve

Your team checks the order beside the email, then it goes to the TMS.

The product · 1 of 3

Every attachment

gets read.

Scans, photos, PDFs, spreadsheets, Word files and forwarded mail, whatever they are called. Shared inboxes and SFTP drops land in one place.

Three cards: intelligent OCR, any file type, every source in one dashboard
An order beside its email: the load site highlighted where it was read, the consignor matched to its TMS code

The product · 2 of 3

Every value

traced and matched.

Each field shows the line it was read from. Customers and sites become your TMS codes, the rest comes from your past orders.

Anything uncertain is flagged for a person, never guessed.

The product · 3 of 3

You approve.

Nothing moves before.

Review the order beside the email, save a draft or approve. Only approval sends the order to your TMS.

The finished order with every field filled and the Approve button

Why it works

Built for real order flows.

Not demos.

A follow-up mail moves the delivery date and the order shows the change

Updates land on the right order

A change later in the thread finds the order it is about.

One mail for two containers becomes two orders, one per movement

One email, several orders

A mail that books several movements becomes one order each.

A correction made on Monday is used by the agent on Tuesday's order

It learns from your team

A correction your team makes is used on the next order.

Where we are

Live.

With a first customer.

First customer live

An intermodal operator runs its order mail through Petaron.

Measured on real orders

Every change to the agent is tested on the customer’s own mails first.

Your data, in the EU

EU-hosted, every customer’s data kept apart, every approval logged.

Petar and Ron, the founders, arms crossed

About us

Petar and Ron.

Founders of Petaron.ai.

Data science and AI, built in Breda. It started as a university project and became a business.

Thank you.

Petaron logo

See it on your own orders: petaron.ai