Why Trace is one AI agent with a team of freight experts

A load is not one job. It is a chain of small ones: a quote to price, a truck to find, a customer to update, an invoice to check. Each link has its own rules, its own counterparties and its own ways to go wrong. A good freight team splits that chain across people who are each good at their part.

We built Trace Agents the same way. It is one agent with a team of experts behind it. You brief the agent. It breaks the task down and hands each part to the expert for it. This post explains why we built it like that, and what it changes for the team that works beside it.

One brief, not six tools

The obvious way to add AI to a freight desk is one tool per task. A quoting tool reads the inbox. A tracking tool makes check calls. A billing tool reads invoices. Each one does its step well and hands the rest back to your team.

The trouble is the hand-offs. The quoting tool does not know which carrier covered the lane last week. The tracking tool does not know that a receiver said on a call that the dock closes at 3 PM. Your team ends up as the glue between tools, carrying context from one screen to the next. That is the work you wanted to hand off in the first place.

One agent removes the glue. You give it a task in plain English, the way you would brief a new hire. The brief is one message and a file, with no forms and no setup screens. The agent works out which experts the task needs and in what order, and it keeps the whole task on one record.

An expert for every part of the load

Behind the agent sit six experts, each with its own procedure:

• Quoting reads requests in any format (emails, PDFs, spreadsheets, screenshots) and prices them from live carrier rates, your lane history and your margin rules.
• Carrier sales posts to DAT, Truckstop, Loadlink and 123Loadboard at once, emails and calls the carriers who run the lane, checks authority, insurance and safety, and ranks every offer by rate, rating and history on the lane.
• Track and trace follows each load through carrier systems, tracking links, ELD data and check calls, and tells customers before they ask.
• Calls works the agent's own phone line: calls to carriers, check calls on loads that go quiet, and the calls that come in, each summarized and filed on the load.
• Claims works from the same load record, so the BOL, the POD and every message about the load are already there, and customers find their claims in the portal beside their PODs and invoices.
• Finance matches carrier invoices to the booking, the BOL and the POD, disputes overcharges with the proof attached, chases receivables, and syncs to QuickBooks or Sage.

Each expert is narrow on purpose. A narrow procedure is easier to check, easier to correct and easier to trust. The agent does the part no single expert can: knowing what the whole task needs, and in what order.

What a task looks like

Here is an example with sample companies. Say a shipper like Aldermoor Supply accepts a quote for a full truckload from Montreal to Boston, load WAL-2482, sold at $2,450. The rep's brief is one line: cover it and keep Aldermoor posted.

The agent hands the posting and the outreach to carrier sales. The load goes to all four boards, and the carriers who have run Montreal to Boston are emailed and called at the same moment, the ones with history on the lane first. Offers come back from the boards, the inbox and the phone, and land in one ranked table, with each carrier's authority, insurance and safety already checked.

In this example, Ridgeway Transport ranks first at $2,050. It is $60 above the lowest offer, and it has been on time on all 12 loads it ran on the lane. The booking and the rate confirmation are prepared with the reasons attached, and they wait for the rep's yes. Once they are approved, track and trace takes over: it shares tracking with Aldermoor once the driver checks in, and chases the carrier if an update goes missing. When the carrier's invoice arrives, finance checks it against the booking, the BOL and the POD.

The rep wrote one line and made one decision. Every step in between sits on the load's record, with the expert that took it.

Why the experts share one record

Experts only help each other if they see what the others learned. That is why every email, call and change lands on the load's record, with who did it and why. A dock hour mentioned on a call is saved to that location, with the call as its source, so the next appointment fits it. A carrier's on-time record shapes the next ranking. When someone on your team corrects a reply, the correction is remembered for the next one.

One brief in, one record out, and an expert behind every step.

The record is also how you stay in charge. While a task runs, the to-dos show each step and what it found. Afterwards, you can see what the agent did, which expert did it and what it used, and correct it where it went wrong.

It works where your team already works

One agent also means one set of connections. It works in Microsoft 365 and Outlook with an inbox of its own, and it has its own phone line for calls and texts. It posts to the four boards, connects to carrier APIs such as XPO, Estes, FedEx Freight, TForce Freight and Day & Ross, and reads and writes the TMS and ERP you run. Your team keeps working in the screens it knows, and nothing has to be replaced to start.

How far the agent goes on its own is yours to set. You pick a mode for each customer and type of freight, from Shadow, where nothing is sent, to God mode, where it works end to end inside your rules. In every mode, bookings and rate confirmations wait for a person, and when the agent is unsure, it asks you with the context attached.

If you want to see one agent run your desk, book a demo and bring a week of your loads.