Orbaion What you could automate
←All work
A freight transport company live

Logistics 2.0: the agent finds the carrier and negotiates the rate

The logistics manager just confirms the finished deal

Logistics 2.0
08:12New request: Novosibirsk → Yekaterinburg, 20 t, curtainsider
08:12Searching the load board… 47 carriers found → 12 after filters
08:32→ negotiating: "Market rate is 72–78. We can pay 74 000"
08:47← "76 000 and a truck tomorrow at 9:00"
08:48company details collected, phone verified → card ready
08:48handed to the logistics manager: confirm the deal ▸ [human]

Re-enactment. Names and figures are illustrative

Problem

Every load takes hours of the logistics manager's time: find a carrier, vet it, negotiate the price down, collect company details. While the manager is busy with that routine, customers get no attention—and the loads keep coming.

What we built

An agent that goes the whole way to a deal on its own: it finds candidates on the freight load board, weeds out the unreliable ones, negotiates in writing and hands the logistics manager a finished card—rate negotiated, company details and phone number verified. The human is left with the decision, not the routine.

How it works

01Request 02Shortlist 03Vetting 04Negotiation 05Confirmation 06Deal
01

Request

The logistics manager enters the load: route, cargo, tonnage. From there the agent works on its own.

02

Shortlist

The agent works through the load board and screens candidates: insurance, rating, age of the firm.

03

Vetting

Questionable carriers drop out before negotiations start—not after the advance is paid.

04

Negotiation

The agent negotiates inside a set rate range—by email and on the load board at the same time, never losing track of who was told what.

05

Confirmation

The carrier's phone and company details are checked before they reach the logistics manager, not after the advance is paid. The agent even pulls the confirmation code out of the inbox without a human.

06 · human

Deal

The logistics manager gets the finished card and confirms it—the final word always belongs to a human.

See the scenario in the Control Room→
Candid

One carrier turned out to be three different legal entities.

The agent learned to tell them apart and to keep three separate cards with the right company details—a mix-up in the paperwork would have cost more than the whole automation did.

Same problem? Message the bot—we start with a review

Email us