The Inbox That Learned to Move Freight

At 9:03 on a Tuesday morning, a freight broker receives an email. There is a PDF attached, perhaps a spreadsheet, and somewhere inside it a truckload must be born. Origin. Destination. Weight. Equipment. Pickup window. A person reads the message, copies the fields into a transportation-management system, consults a pricing engine, writes back, phones a carrier, bargains, books a dock, and later asks a driver where the truck is. The work is essential. It is also a splendid way to turn a capable operator into a human keyboard.

Vooma's proposition is that the software should do the keyboard work and the operator should handle the judgment. The San Francisco company builds freight-specific AI co-workers that read, call, text, log into portals and update the systems already on the desk.

The narrow door

Jesse Buckingham and Mike Carter chose an unfashionably small entrance. Buckingham had run a group of logistics-software businesses, growing annual recurring revenue from roughly $2 million to more than $20 million. He knew the customers and their recurring irritations. Carter had been employee number two and a founding engineer at Kodiak Robotics, after work on autonomy at Otto and Uber ATG. He knew how to make software operate in a physical world that does not politely follow the happy path.

They met through a mutual friend and launched Vooma through Y Combinator's Winter 2023 batch. The original product automated order-taking from shippers. Their launch argument was blunt: about 60% of orders arrived as email, text, spreadsheets and PDFs, then somebody keyed them in by hand. Before one could automate trucking's grand choreography, one first had to stop retyping the attachment.

That pairing explains the product better than the phrase “AI agent” does. Vooma is neither a new TMS nor a handsome dashboard asking for another login. It sits across the systems a freight team already uses. Quote reads the email, calls the pricing tools and drafts the reply. Build extracts the order and writes it into the TMS. Schedule negotiates with a facility through email, phone or portal. Cover posts the load, answers carriers, checks credentials and records bids. Track makes the check calls. Collect hunts for the proof of delivery.

The first thing to fail was the patch

The revealing customer stories begin not with failed logistics, but failed human patches. Direct Traffic Solutions had installed a modern TMS and phone system. Still, a lean team answered only 39% of thousands of inbound carrier-sales calls. The unanswered majority went to voicemail, taking rates, offers and possible capacity with it. Managers also lost the market intelligence contained in the calls they never heard.

The company was wary of voice automation. Freight calls are short, peculiar and suspicious by nature; a carrier who meets a dim robot simply calls the next broker. A month-long pilot changed the calculation. Vooma Cover answered and logged the calls, qualified carriers and sent the promising conversations to people. Direct Traffic Solutions says reps went from about five booked loads a day to ten, while Vooma-booked loads produced 5% higher margins. The quiet surprise was not the voice. It was the new record of more than 1,000 carrier offers each week.

“We knew we were missing calls, but we didn't realize the opportunity cost until we saw the numbers.”Chris Griffin, Director of Safety, Direct Traffic Solutions

Whitewater Freight arrived at the same answer through headcount. Its president, William Bochkay, calculated that keeping up with carrier calls meant hiring three additional reps. He also feared a fully automated caller would sand away the relationships his company had spent decades earning. Six weeks after adopting Vooma, staff-handled calls had fallen from 2,900 to 1,150 a week, with 838 offers recorded and 86% of posted loads booked. The humans did not disappear. The indiscriminate ringing did.

The product is the handoff

Plenty of companies now sell AI to freight. HappyRobot is prominent in voice; Augment markets a broader AI teammate; FleetWorks, CloneOps, Pallet, Parade and Numeo attack adjacent pieces. TMS vendors are adding agents of their own. Vooma's distinction is not possession of a language model. Models are rented by the hour. Its argument is operational breadth joined to freight detail: one agent can move from an unstructured email to a rate engine, from a carrier call to a TMS entry, and from a tracking exception to a Teams message.

The company's “Memories” feature gives away the hard part. Every freight business keeps a private book of lore: never quote this customer without liftgate fees; that warehouse dislikes early arrivals; this lane needs a different markup on Fridays. Vooma lets teams express those rules in natural language and choose how far automation goes - click-to-quote, auto-draft or auto-reply. The real product is not synthetic speech. It is the controlled handoff between machine action, company memory and human exception.

This is also why Vooma integrates with McLeod, Turvo and 3PL Systems; rate tools such as DAT, Greenscreens and Truckstop; carrier-vetting service Highway; collaboration tools including Slack and Teams; and scheduling portals from Opendock to Blue Yonder. A freestanding intelligence is merely clever. An intelligence that can write the result back is an employee.

The economics of saved attention

Vooma sells enterprise software through demonstrations and tailored deployments. The buyer's arithmetic is less mysterious than the price sheet. Makt-Trans estimates it avoids at least $50,000 a year in additional order-entry or account-support staffing. Its contract-rate entry fell from five minutes to 15 seconds, quote replies from eight minutes to two, and weekly sales calls rose by more than half. Evans Transportation says 90% of orders are now processed automatically, representing more than six reps' worth of manual load-building time each year.

The company itself has followed the same expansion logic. It reported 12.5-fold revenue growth and a 32-fold increase in transaction volume in the year before its funding announcement. A $3.6 million seed round led by Index Ventures and a $13 million Series A led by Craft Ventures brought disclosed funding to $16.6 million. The money went toward engineering, sales, support and the broader Agents and Voice products. Investors included a conspicuous number of logistics operators and founders - useful patrons in a market where the API documentation is sometimes a veteran's memory.

The fine print is the work

There is a copyable lesson here, and it is smaller than “buy AI.” Start with a high-volume task whose beginning and end can be counted. Write down the operating procedure. Connect the systems of record. Decide which exceptions deserve a person. Run a pilot long enough to observe not only labor saved but information newly captured. Direct Traffic Solutions learned that declined calls were pricing data. Sunset learned that quote logs could expose changes in customer volume. Automation made the invisible work measurable.

The conditions matter. A company with low volume, inaccessible systems or undocumented rules will struggle to recover the integration cost. A team that changes nothing about roles may simply bolt an agent onto the same old queue. Voice automation that cannot transfer a delicate call will damage the relationship it was meant to protect. And 90% accuracy can be dreadful when the remaining 10% contains a missed pickup, fraudulent carrier or six-figure load.

Vooma's better customer stories preserve a human boundary. Carrier reps stop answering callers they cannot use and spend time sourcing capacity. Track-and-trace teams stop asking every driver the same question and investigate the 5% that looks wrong. The machine supplies consistency and concurrency; the person supplies suspicion, tact and improvisation. That arrangement is less cinematic than an autonomous supply chain. It is also how work actually changes.

The company began by removing clicks from a freight inbox. Its larger ambition is to remove the waiting between one person, one system and the next.

In November 2025, Vooma recast its range as “AI co-workers” moving freight from quote to cash. By mid-2026 it was pushing Track, where agents call every carrier at the prescribed time, write milestone updates into the TMS and leave the morning shift a list of genuine exceptions. It is a grander story now. Yet the original insight survives: freight does not need to become orderly before software can help it. The software must learn where the disorder lives.