He spent four years inside Uber Freight watching brokers chase loads by phone. Then he built the machine that makes the calls.
Paul Singer's product does something deceptively dull: it picks up the phone. In American freight, a load gets booked, a driver gets a check call, a rate gets haggled - and almost all of it happens by voice and email, person to person, thousands of times an hour. FleetWorks, the company Singer runs, built voice AI agents that handle exactly that grind. One agent works for carriers, learning their trucks, lanes and availability. Another works for brokers, hunting the best available truck for every shipment. The agents do not sleep, do not miss a check call, and do not get stuck on hold.
That is the pitch, and it lands because Singer is not guessing about the problem. He lived it. For roughly four years at Uber Freight, he sat in carrier product - the side of the business that deals with the men and women who actually move the trucks. He watched how much of a supposedly digitized marketplace still depended on a rep dialing a number and asking, again, where the driver was. When he left in December 2022 to start FleetWorks with cofounder Quang Tran, he was not chasing a trend. He was going back to fix the thing he had stared at for years.
For a $1T industry, I'm always reminded of how much still runs on phone calls and email.
FleetWorks went through Y Combinator's Summer 2023 batch and moved fast. Within the first six months, Singer says the company brought more than 10,000 carriers and dozens of brokers onto its platform - including, in a detail he clearly enjoys, his old employer Uber Freight. The product matured from matching trucks with cargo into an "always-on" AI dispatcher that can run the repetitive carrier-and-broker conversations end to end. Singer frames the value in the only language freight brokers care about: he claims the system returns about 10x ROI in operating costs for broker customers.
The freight world is crowded with people promising AI will change everything. Singer's edge is specificity. He is not trying to replace the broker; he is trying to delete the phone tag. The agents learn preferences, equipment and quirks the way a good carrier rep would, then handle the volume no human team can cover around the clock. It is the unsexy middle of the supply chain, and that is precisely the point - the boring part is where the money leaks out.
In October 2025, FleetWorks announced $17 million in funding, anchored by a $15 million Series A led by Bill Trenchard at First Round Capital. The symmetry is hard to miss: Trenchard also led Uber's seed round back in 2010, years before Singer would join Uber Freight and years before he would leave it. Y Combinator, Saga Ventures and LFX Venture Partners joined the round. The capital is going where you would expect a fast company to spend it - hiring, commercial expansion, and pushing the AI dispatcher further into production.
On why he picked First Round, Singer was blunt about pedigree. The firm, he said, is "the preeminent investor in early-stage investing." It is the kind of line that sounds like flattery until you remember the through-line: the same investor, the same conviction about a logistics bet, fifteen years apart.
We ultimately went with First Round Capital to lead the round because they are the preeminent investor in early-stage investing.
Before freight, there was Yale. Singer graduated in 2016 with a B.A. in economics, and filled out the edges of his degree with Chinese language and mechanical engineering - an unusual combination for someone who would end up automating dispatch calls, or maybe exactly the right one. His early resume reads like a finance-track economics major: an associate role at L.E.K. Consulting building market models, a summer analyst stint at Point72 Asset Management, and project work with Design for America that put him alongside FEMA on communications problems.
Then in 2018 he answered a job posting for a Strategy & Planning Associate at Uber and landed in Uber Freight, where he spent the next four-plus years in product and strategy roles on the carrier side. That is the apprenticeship that matters here. Singer did not parachute into trucking with a deck and a thesis. He learned the workflow from inside one of the most ambitious attempts to digitize it, then concluded the real unlock was the part everyone treated as too mundane to fix.
His cofounder Quang Tran brings a different flavor of ambition - he worked on "moonshot projects" at Airbnb. Together they aimed at a target that is anything but moonshot in feel and entirely moonshot in scale: the roughly 200,000 truckload transactions that happen in the US every single day.
Singer wears the founder role with a dry sense of humor. He has joked about interviewing engineers who had never seen the movie Shrek, admitting he never expected to feel like an elder in the tech industry. It is a small thing, but it tells you something: he is building a serious company in a deeply unglamorous corner of the economy, and he is not pretending the work is anything other than what it is.
"For a $1T industry, I'm always reminded of how much still runs on phone calls and email."