Stakh Vozniak and the problem of the second phone call

Stakh Vozniak made up to sixty calls a day as a dispatcher for his own startup. While he was talking about one load, other loads were disappearing. He was trying to build a company that would make finding freight easier, and the telephone kept explaining why that company needed to exist. Every conversation occupied the same scarce resource: his attention. A truck could be ready, a customer could be willing, and the person connecting them could still be busy on another call.

That is a useful place to begin with the co-founder and CEO of Cargofy. Today he sells AI workers that can call carriers, collect bids and follow shipments. Before that, he did the work himself. The distance between those two jobs runs through a Ukrainian juice factory, an Israeli accelerator, Cedar Rapids and several ideas that customers politely declined to love.

The apples came before the algorithm

Vozniak was born in Kalynivka Druha, in central Ukraine. Entrepreneurship arrived early: at thirteen, he was selling DVDs. The first business was small enough to fit into a schoolboy’s life. The impulse to find something people would pay for has remained rather more durable than the format.

Computers offered another route. At school, he worked nights in a computer club to get access to a machine. He saved for his own, learned programming and developed games for a distributor. Two were released. A third needed a larger team than he had allowed for, consumed the money from its predecessors and never reached launch.

Then came apples. A juice factory needed fruit moved during the busy season, and finding enough trucks was a problem. Vozniak helped connect loads with drivers using text messages. Software, in this setting, had an unusually plain test: did the apples get where they needed to go?

He built Wozward, a conventional freight business, before making another attempt to put technology closer to the customer. Tona, developed in the early 2010s, offered on-demand urban freight. The experience gave him something harder to acquire than an appealing product description: familiarity with what happens when an actual shipment needs an actual vehicle.

An education at the truck stop

His geographical ambition expanded before the American move. In 2016, Cargofy participated in Capsula.Studio, an Israeli transportation accelerator. The program helped him look beyond Ukraine and think about a global business. In 2017 he applied to three accelerators and won a place in the Iowa Startup Accelerator.

Cedar Rapids was a practical setting for a freight company. Iowa offered proximity to major transport routes and the Iowa 80 truck stop, along with lower operating costs than California. The American company maintained a research and development presence in Kyiv. The arrangement joined local market learning to a team back home, rather than requiring every piece of the business to occupy the same address.

He arrived in Cedar Rapids with his wife, Nataliia, who was then Cargofy’s chief operating officer. By that point he had visited thirty-three countries. Travel had supplied breadth; Iowa required depth. He needed to understand how a particular kind of American business owner earned a living.

The early customer was the owner-operator, a driver with a truck and a business to run. Freight boards list available loads, but a list cannot make the decision, complete the paperwork or negotiate the arrangement by itself. A person doing all of that has less time available for the activity that pays: moving freight.

“Owner-operators want to make money without being the dispatcher.”

Stakh Vozniak, 2017

Vozniak’s early explanation of Cargofy was a virtual assistant for truckers, using a familiar phone assistant as shorthand. It could match loads, help plan routes and track shipments. The founders included Alex Kovalchuk on technology and Dimitris Alexiou on the commercial side. Each brought a different part of the problem into the room.

The gold stars nobody ordered

Cargofy initially concentrated on Eastern European immigrants driving in the United States. Localized software and support in their own languages gave those customers a way into a complicated market. For a founder who had moved countries himself, the difficulties were close at hand. A new arrival needed useful freight and help navigating the surrounding business chores.

Vozniak also tried adding digital rewards for completed trips. Drivers were unimpressed. A built-in navigation product went wrong too: development costs were followed by a jump in the price of the underlying software kit. The project was abandoned. These were expensive ways to discover that a feature can please its inventor considerably more than its intended user.

He has described spending too long polishing the interface. “Customers didn’t care.” The sentence has the advantage of fitting on a very small product brief. His own dispatch calls gave him better material: the work was slow, competing opportunities vanished, and the bottleneck persisted regardless of how handsome the screen looked.

There is an appealing awkwardness here. Someone who began by making games had tried to give truckers rewards for doing their jobs. Truckers already had a scoring system. It arrived as money. The episode makes his later emphasis on practical operations feel earned.

When the assistant could finally take the call

The company’s AI ambitions preceded the technology’s usefulness. Cargofy dates its commitment to AI freight operations to 2018, while its timeline places the first production AI workers in 2023. The intervening years were spent building the carrier network and operating the business. That sequence matters: the freight experience came before a language model could converse convincingly about it.

The current product is broader than the original driver app. Companies can assign digital workers to freight procurement, documents, tracking, quoting and finance. Cargofy describes a worker as having a role and permissions. It connects those workers to the calls, emails, WhatsApp messages and systems already present in the customer’s day.

The appeal is easy to picture without imagining a driverless motorway. A shipment generates a series of small administrative obligations. Someone must ask for prices, compare replies, check progress and collect the delivery documents. Completing those obligations requires repeated contact with other people. A dashboard can display the unfinished business very neatly while leaving it unfinished.

Vozniak’s early complaint about the telephone now has a technical answer. Several conversations can progress while a human handles an exception. The test remains the old one: whether the work gets done. The apples have acquired more elaborate paperwork.

Three founders, and a larger room

Cargofy’s founding team combines distinct operating histories. Kovalchuk, the CTO, trained in telecommunications and radio engineering and worked on spacecraft launch systems. Alexiou brought experience as a freight broker and trader at Maersk and Mitsui OSK Lines. Vozniak brought the businesses he had already run and the customers he had spent time trying to understand.

In June 2026, Cargofy announced its Series A. The widely reported $11 million total comprised $6 million in primary capital and $5 million in secondary transactions. The distinction is consequential: new investment finances the company; secondary transactions buy existing shares. Putting both beneath one headline can make a company’s available spending money look larger than it is.

Investors included u.ventures, Toloka and Movens Capital, with participation from Intercom co-founder Des Traynor. Movens reported more than 2,000 paying customers and over $10 million in annualized revenue. Those are company-scale figures, rather than a measure of the founder’s personal wealth.

In July, Vozniak, Kovalchuk and Alexiou became Endeavor Entrepreneurs following selection in Athens. By then the customer map extended across North America, Europe and Central Asia. The accelerator applicant who once needed an introduction to American trucking was working inside a much larger network.

Trust still has a human job

Giving software a job title does not automatically persuade a customer to trust it. Vozniak has described an experiment with agents named after people customers already knew, rather than unfamiliar fictional colleagues. The company found that this changed how customers received the technology. Familiarity turned out to be an operating consideration.

His examples of adoption are often modest and specific. One fleet wanted calls to drivers who left engines running at truck stops. The question was practical: could the system notice a defined event, apply a company rule and contact the right person? Such assignments offer a less theatrical introduction to AI than a promise to remake an entire business before lunch.

He also acknowledges failed deployments and the need for customer participation. The company has to understand a process before it can automate it. That requirement returns him to the work he did in Iowa: learn what people actually do, including the inconvenient parts that never appear in a polished demonstration.

A global company with something to bring home

His recent calendar reflects the next stage. In September 2026 he described a week at Stanford’s Innovation and Growth Program with Endeavor Ukraine, studying pricing, governance, disruption and venture capital. In October he appeared in the opening episode of Laba’s business podcast, discussing Cargofy’s American expansion and investment.

Away from freight, his personal website resembles a computer desktop. There are movable application windows, a falling-block game and a conversational StakhGPT. The schoolboy who wanted access to a computer remains visible in the interface. A conventional executive biography would have used considerably fewer buttons.

His longer ambition has a geographical dimension. He wants a global business whose eventual success leaves knowledge, experienced managers and some of its financial rewards in Ukraine. An acquisition or a public listing would matter partly for what people could build afterward. He has studied the example of countries where one generation of founders supplies the next.

The sixty calls help explain the continuity. Vozniak keeps returning to work that consumes somebody’s day, then trying to build a way through it. The setting has changed, the financing has grown, and the software can now answer back. There is still a load waiting for somebody to pick up.