Fura says its AI agent books over 40% of freight loads; a lead signs off

Fura, a freight broker with more than $100 million in revenue, says its AI agent, Donna, now handles more than 40% of the company's load bookings: finding a truck for a shipper's load and agreeing the rate, work a carrier sales rep does by phone. Chief executive Jeff Dangelo said so in a video interview FreightWaves published on 11 November 2026, and, according to FreightWaves's summary, expects 90% by the end of next year.

The summary says Donna makes and receives phone calls, negotiates rates and books the load. It is connected to Fura's pricing algorithm, its transportation management system (the software that tracks each load) and Highway, which Fura uses for carrier compliance, the checks on a carrier before it is hired. One step stays with a person.

Donna is doing all of that work from compliance to negotiating to booking. We do have her handing off the finished sort of negotiation to a team lead.

Jeff Dangelo, chief executive of Fura, as quoted by FreightWaves, 11 November 2026

What the 40% does and does not measure

• No definition. The summary does not say what counts as a booking "handled" by the agent, over what period, or out of how many loads.
• No clean baseline. In August, FreightWaves's summary of an earlier interview said "nearly 40% of booked carriers are now sourced through AI agents operating via email, text, and phone". Finding a carrier is a narrower job than booking a load start to finish, so the two figures are not a before and after.
• The 90% is a forecast with two dates. The headline and summary say the end of next year; in the quoted words Dangelo says "even by the end of the year getting closer and closer to 90%". He calls 100% a longer-term goal, because each company Fura buys adds routes and carrier relationships that take time to fold in.
• No error rate. Nothing says how often the team lead rejects a deal the agent negotiated, or whether an agent-booked carrier has failed a check afterwards.

The people side

Fura grows by buying small brokerages and moving them onto its software; the latest, Road Dog Logistics, was its eighth, FreightWaves reported on 20 October. In a June interview summary, Dangelo said headcount at one acquired brokerage, Pinwheel, went from 26 to 8 while it moved from a $150,000 loss to a $1 million profit. The same summary credits offshore staff and a standard operating model as well as automation, so the headcount fall cannot be put down to AI alone.

Dangelo put Fura's overhead (selling, general and administrative cost, or SG&A) at roughly 39% in June, against 60% to 80% for the industry by his account; the summaries do not say 39% of what. This week he said Fura makes $2.5 million in revenue per employee, and that it is adding to its carrier sales and carrier success teams over the next three to six months because trucks are getting harder to find. None of the pieces says what happened to the people who left acquired brokerages.

The catch

Every figure is the chief executive's, with no method, and none of the four pieces gives the cost of building or running the agent, or says whether Fura built it or bought it. The stakes of a wrong booking have risen: FreightWaves's summary says the US Supreme Court ruled 9-0, in a case it calls Montgomery, that freight brokers can be sued under state law for hiring unsafe carriers. Fura says it checks every shipment by software, with a person giving the final sign-off.

What to do if you run carrier sales or logistics operations

• Before you trial a voice agent on carrier calls, write down which step a named person signs. Fura keeps two: a team lead confirms the negotiated deal and a person gives final sign-off on carrier compliance.
• When a vendor quotes a share of loads "booked by AI", ask whether that means booked start to finish or carriers found, over what period, and how often a person overturned the agent.
• Ask your insurer or counsel who is liable when an agent books a carrier that turns out to be unsafe, before the agent makes its first live call.