FleetWorks

Somewhere, a truck is empty and a phone is ringing.

It is a Tuesday at a mid-sized freight brokerage, and the load board is screaming. A reefer needs to leave Fresno tonight. Somewhere out there is a driver with the right equipment, in the right place, willing to take it at the right price. Finding that driver used to mean a human dialing through a phone tree, leaving a voicemail, sending a text, refreshing an email, and starting over. Dozens of times. Per load.

At FleetWorks' customers, that work increasingly happens without a human dialing at all. An AI agent picks up, places the call, navigates the menu, asks the carrier where they're empty, confirms the equipment, and books the load. It does this thousands of times a day, in parallel, on phone, text, and email - and it does not take lunch. FleetWorks is a San Francisco software company that has, in remarkably short order, convinced the freight industry that the most repetitive job in logistics can be handed to a machine.

That single sentence is the entire pitch. Freight is a $1 trillion market that still runs on roughly a billion phone calls a year. FleetWorks looked at that number and saw not a tragedy but a to-do list.

The system for matching trucks with freight is slow, manual, and gloriously fragmented.

Two operators who had seen the inside of the machine.

FleetWorks was founded in 2023 by Paul Singer and Quang Tran, and the resumes are not incidental. Singer was a product manager at Uber Freight, where he helped build some of the industry's first real-time autonomous systems for managing carrier quality. He had watched, up close, how much of freight still ran on manual phone work - and how badly it scaled.

Tran came from Airbnb, where he worked on what the company politely calls "moonshot projects." Between them they had the two halves of the problem: one founder who knew freight's plumbing intimately, and one who knew how to build ambitious systems that talk to the real world. They took the idea through Y Combinator's Summer 2023 batch and started, deliberately, on the broker side of the market.

An always-on dispatcher that happens to be software.

FleetWorks started as a tool for brokers: an AI carrier-sales rep that automates the outbound and inbound calls, texts, and emails needed to cover a load, plugged directly into the brokerage's transportation management system. The pitch to a broker is blunt and measurable - one customer reported a single rep booking 50 to 60 loads a day, a number that used to belong to a small team.

In October 2025, alongside its Series A, the company opened the other side of the market. It launched an "always-on dispatcher" aimed at carriers themselves. This agent works with thousands of carriers every day - over phone, text, and email - to learn what equipment they run, where they are, and when they'll be empty, then matches them to loads while accounting for the fussy details that matter: facility requirements, driver preferences, the things a spreadsheet ignores.

The technology underneath is voice synthesis and generative AI tuned to each carrier's preferred channel. The unglamorous-but-telling detail: the system navigates automated phone trees - the menus everyone hates - with a 96% success rate. It can also do things humans can't easily do at scale, like voice verification to fight the freight industry's persistent fraud problem.

The result is a two-sided marketplace where the matching, the calling, and the booking are increasingly automated - and the humans are pushed up the value chain toward the work that actually needs a human.

Investors who backed Uber and Flexport are paying attention.

Money is a lagging indicator, but it is an indicator. FleetWorks' $17M Series A was led by First Round Capital, with partner Bill Trenchard taking the lead - the same investor who led Uber's 2010 seed round and backed Flexport early. Y Combinator, Saga Ventures, and LFX Venture Partners joined, bringing the company's total funding to roughly $19 million.

Trenchard's thesis is refreshingly anti-software-marketing: he argues that the old playbook simply doesn't fit this job. The adoption backs it up. More than 40 U.S. brokerages now run FleetWorks, among them 15 of the top 100 by volume - and, in a detail too good to invent, Uber Freight, the company the CEO left to start this one.

Not replacing the rep. Promoting them.

It would be easy to read FleetWorks as an automation story with a familiar ending - software arrives, jobs leave. Singer pointedly tells a different one. The humans, he argues, are the powerful part of the process: they build relationships and untangle the messy edge cases that no model handles cleanly. The repetitive dialing is what the AI is for.

His framing is that AI will not take the carrier-sales job - it will change what that job looks like. The rep who used to spend the day on hold becomes someone who closes the hard deals and watches a fleet of agents handle the volume. Whether every brokerage reads it that generously is a separate question. But as a stated mission, it is a specific bet: that the winners will be the brokers who get good at using the tools, not the ones who resist them.

Back to that empty truck in Fresno.

Return to where we started: a Tuesday, a reefer that needs to leave tonight, a load board screaming. In the old version of this scene, a tired human starts dialing and hopes. In the FleetWorks version, the call has already been placed, the carrier already asked, the equipment already confirmed - while the human was busy with something only a human can do.

That is the whole argument, and it is smaller and more concrete than most AI pitches. Freight doesn't need to be reinvented. It needs the billion annual phone calls to stop being a person's whole day. If FleetWorks is right, the most overworked employee in logistics over the next decade won't be a dispatcher with a headset. It'll be software that never puts the phone down - and a human who finally got to hang theirs up.

The trucks were always ready. FleetWorks is betting the conversation can finally keep up.