The checkout button looked like progress. Tamas Kadar and Bence Jendruszak were university students with a cryptocurrency exchange and an apparently clever proposition: skip the lumbering bank transfer, charge a card and deliver the crypto immediately. Customers could avoid waiting three or four days. The internet, as it often does, supplied a brutal product review. Fraudsters found them almost at once. Within the first week, Kadar says, half the venture's revenue was gone - a few thousand dollars swept away through a problem the founders barely knew had a name.
They did not yet understand chargebacks, stolen payment instruments or the small industry devoted to defeating both. So they gave their summer to the criminals. The two students read darknet forums, studied schemes and tried to understand where personal information came from and how it was put to work. Available fraud software was built for enterprises, priced and packaged for companies much larger than theirs. Worse, it missed some of the signals they thought mattered. They began writing their own defense.
The irony is almost too neat: the crypto business that failed to protect itself produced a fraud product other businesses wanted. Exchanges began asking to use it. In 2017, the side solution became the company. They called it SEON.
A partnership built from opposite instincts
Kadar met Jendruszak at Corvinus University of Budapest in 2013. Cryptocurrency was the first common language. Their temperaments were not. Kadar describes himself as the loud one, capable of talking for hours, inclined to jump into an API document while imagining how a new product might look and whom he might sell it to. Jendruszak goes to the practical questions: Do they need to hire? Where will those people come from? What will it cost? What happens to the cash?
Their executive team calls the pair Yin and Yang, which is office shorthand for a useful tension. Kadar thinks six months ahead and makes the destination vivid. Jendruszak orders the steps. In a culture that likes to turn founders into singular heroes, their arrangement is a reminder that vision improves when somebody nearby is prepared to ask about the invoice.
“Find collaborators you can trust. Nobody starts a great business alone.”Tamas Kadar
The trust is old enough to contain disagreement. Kadar has written that nothing important between them is left unsaid. He was the friend Jendruszak called after breakups; years later, when Jendruszak was at the hospital for the birth of his second child, Kadar spent the night looking after his older son. Their conversations now range from family logistics to deepfake identity documents. The scale changed. The candor did not.
The cheapest office furniture in Budapest
SEON's first year was financed with €50,000, second-hand furniture and whatever cloud credits Kadar could scavenge. He messaged people on Skype asking for AWS promo codes. The team appeared at events that handed out credits. This was less startup folklore than a weekly calculation: every expense avoided bought more time to find product-market fit. The savings helped the founders hire two university students to build out the product.
Nine years and $187 million in total funding later, Kadar still tells the story as a lesson about signals. How founders spend teaches the company what matters. Loose money spent on work that does not create growth or revenue quietly grants everyone else permission to do the same. Scarcity made the lesson impossible to miss. Abundance merely makes it easier to forget.
From dorm-room fix to global platform
There was another early lesson, absorbed before SEON existed. At 15, Kadar ran a small business reselling smartphones that he flashed into Hungarian. He loved the technical work and bristled at customers who asked too many questions. His parents had both freelanced - his mother as an accountant, his father as an electrical engineer - and his father offered a correction. The customer pays. Be kind, and the reputation will travel.
Kadar eventually turned the admonition into a rather extreme management habit: for SEON's first six years, he reviewed every customer-support ticket. The queue showed him where clients were confused, but also what the company ought to build. The founder who once wished customers would simply understand learned to regard their questions as unpaid product research.
Listen to the question. Find the repeated friction. Turn it into a product decision. Then watch the queue again.
Security without the interrogation room
Fraud prevention has a built-in comic problem: its perfect customer experience would be invisible, while its failures are spectacular. Block too little and the company loses money. Block too much and the customer, now presumed guilty by a form, abandons the purchase. Kadar's product argument lives between those outcomes. An email address, phone number, IP address and device can generate a richer digital footprint than a business might expect. The aim is to reserve intrusive checks for moments that deserve them.
SEON grew around that idea. Its software combines digital-footprint data, device intelligence, rules and machine-learning models to help companies decide whether an account or transaction looks suspicious. By 2022, Kadar reported more than $900,000 in monthly revenue, around 200 direct customers and more than 90 million API calls in a single month. The company later expanded beyond point fraud checks into identity verification, transaction monitoring, customer screening and case management.
The ambition now is a connected command center for fraud prevention and anti-money-laundering work. Kadar argues that risk often hides in the handoffs. One team owns onboarding, another payments, another regulatory reporting. A person who passed an identity check months ago can return through an unfamiliar device, and no single team sees the entire journey. His answer is to put those events on one timeline and make decisions in real time.
That ambition helped attract an $80 million Series C led by Sixth Street Growth in September 2025. The round brought SEON's reported total funding to $187 million and funded expansion in North America, Asia-Pacific and Latin America. The company now operates from Austin, London, Budapest and Singapore. Kadar, born and raised in Hungary, has lived in Malta and London and is now based in Austin, though he says he still divides his time between Europe and the United States.
When code gets fast, judgment gets expensive
Kadar's public persona remains closer to product lead than ceremonial chief executive. He writes about synthetic identities, account takeovers and the odd economics of fraud. In one recent example, he said he found his co-founder's Social Security number for sale on Telegram for four dollars, only months after Jendruszak moved to the United States. His point was not the bargain. It was that a permanent nine-digit credential behaves like an old magnetic stripe: once copied, it cannot defend itself.
He is equally wary of using artificial intelligence as decorative upholstery. Faster models make code generation easier, but a fraud system cannot casually ship a change that blocks legitimate buyers or opens a gap for criminals. In Kadar's formulation, quality assurance and judgment become the bottleneck. The barrier to producing output falls; the standard for releasing it must rise.
“Constraints lead to creativity. When resources are finite, you're forced to find a way.”Tamas Kadar
His leadership language returns repeatedly to autonomy. Micromanagement is impractical across offices and time zones, and, he says, uninteresting anyway. Give talented people the priorities and support, then allow them to surprise you. The anti-hierarchical ideal produced a favorite recent moment: a new employee approached Kadar on his first day and asked who he was. Kadar said he was one of the data guys. He was delighted that the answer seemed plausible.
There is a neat continuity here. The teenager who preferred phones to their owners became the chief executive who read the support queue. The student who knew nothing about chargebacks became a fraud specialist by examining the loss instead of merely resenting it. The founder who once hunted free server credits now has institutional investors and offices on four continents, but still talks about spending as an act of instruction.
Kadar once compressed his life into three words: “Relentless problem-solving.” The phrase risks sounding like a slogan until one notices how literal it has been. A checkout failed. He built a filter. Customers complained. He read the tickets. Fraud moved between departments. He set out to join the timeline. The problems keep changing costumes, which is the fraudster's principal talent. Kadar's answer is to keep watching the seams.