Andrej Bencic and the value of a second try

Tenderly began with the sort of assignment that makes a hackathon sound like an earnest weekend: improve public procurement. Build a way for governments to receive bids without requiring everyone to trust everyone else. Andrej Bencic and his collaborators made the platform. Then they discovered a more immediate customer for their ingenuity: themselves.

The equipment for building on Ethereum was frustratingly sparse. The team could see the possibilities in smart contracts, but working out what those contracts were actually doing took effort that familiar software tools had already taught them to resent. Procurement also meant dealing with government buyers. Developer tools meant solving a problem they understood from the inside.

They changed the business and kept the name. A tender became Tenderly, a small linguistic souvenir of a destination they had decided against. Years later, Bencic would lead a company selling ways to inspect, test and simulate blockchain activity. The first useful experiment had been the company itself: try an idea, examine the result, change course before the commitment becomes permanent.

Friends, deadlines and a tolerable amount of chaos

Bencic and fellow co-founder Bogdan Habic met in high school in Serbia. By 2019, Bencic said they had worked together for about six years, across four companies, and had started two together. Tenderly grew out of a working relationship with a history, rather than an introduction arranged for the occasion of founding a startup.

Their competitions produced a trail of experiments. In 2015, Bencic helped build Spark, a nightlife heat map that processed time and location information from Twitter and Instagram activity. It placed second at a 24-hour hackathon in Belgrade. The project let people replay activity across a city. Even a night out, apparently, could benefit from better observability.

By March 2019, the subject had narrowed to transactions. TxFlow, a first-place project at ETHParis, displayed the execution of an Ethereum transaction on a page, including calls between contracts and the variables at a selected point. Bencic was working on a recurring question: how do you make a complicated sequence legible to the person responsible for it?

The early company divided that question among different specialties. In February 2019, Tenderly’s then product lead Vukašin Vukoje described Bencic as handling infrastructure and backend work. Nebojša Urošević concentrated on blockchain and low-level problems, Miljan Tekić mainly on frontend work, and Habic covered a broad remainder. The eventual CEO was building underneath the interface.

That background matters. Bencic had worked on large distributed cloud systems and spent time at MVP Workshop in 2018 and 2019. His point of comparison was a software world where helpful services already existed. He wanted developers arriving in blockchain to have some of that same practical assistance, instead of having to manufacture every missing instrument themselves.

A computer inside the computer

At ETHBerlin in August 2019, Bencic and Habic took the idea of building their own instrument rather literally. Their winning project, sol.tty, implemented an Ethereum Virtual Machine in Solidity, the language used to write programs for that virtual machine. It was a machine built inside the machine, with the modest weekend aim of making experimentation less cumbersome.

The pair spent about 40 hours working through a table of Ethereum instructions. The result offered a REPL-like environment: a place to try small pieces of code and inspect what happened. The name combined Solidity with the Unix term TTY. Bencic explained the pronunciation as “solty”. A pun, at least, required no further infrastructure.

The appeal of that project is easy to understand without knowing Solidity. Each extra deployment and cleanup step interrupts an experiment. A tool that shortens that loop gives its user more attempts. Hackathons had rewarded the team for completing projects under a clock; Tenderly gave Bencic a reason to work on the clock itself.

“We really enjoyed building dev tooling.”Andrej Bencic · 2019

When the weekend project needs a payroll

By July 2021, Bencic was announcing a $15.3 million Series A led by Accel. It followed a $3.3 million seed round earlier that year. The investors included people who had built developer businesses themselves, among them Algolia co-founder Nicolas Dessaigne, Instana co-founder Mirko Novakovic and Vercel CEO Guillermo Rauch. Their experience was part of the attraction.

Bencic’s account of the raise gave users a substantial share of the credit. He said their help had shaped more than half of Tenderly. That is a revealing way for an engineer running a company to describe ownership of its ideas. The product was developing through use, questions and feedback, with customers helping identify what deserved to exist next.

The transition also changed what the job demanded. His February 2022 Chair interview ranged from acquiring customers to the effect of investment and the difference between early challenges and later ones. The conversation’s itinerary tells its own small story. A founder who begins by removing technical obstacles eventually acquires a calendar full of organizational ones.

On March 2, 2022, he announced a $40 million Series B led by Spark Capital. Yasmin Razavi became another important investor relationship. Tenderly had grown from seven people in Belgrade to a team of 37 across continents. Those are dated figures from the announcement, a snapshot of the moment an intimate engineering operation began taking a different shape.

The geography changed along with the headcount. By May 2022, Bencic was describing his own first experience of remote work as the business established a US office. His ambition was to combine strengths from Serbia and the United States. There is a difference between building with people you have known for years and building an organization that people you have never met can join.

A garage with room for other people

Tenderly also opened its circle in Belgrade. In March 2022, the company launched Tenderly Garage, an online and offline initiative for startup education and connections in Serbia. Bencic described it as a way to return support that had mattered while the company was being built. The beneficiaries would include people outside Tenderly’s own payroll.

Asked whether Tenderly might become Serbia’s first unicorn, he moved the conversation toward the wider rise in investment in Serbian startups and companies with Serbian teams. A valuation milestone was less interesting to him than evidence that the local community was maturing. It was a practical answer to a question designed to produce a celebratory headline.

There is a recognizable continuity here. First, a few friends build tools because they need them. Later, customers influence what those tools become. Then the company creates a place for other founders to exchange knowledge. The circle gets larger, while the useful unit remains fairly small: someone has a problem, and someone nearby knows a way through it.

Give the experiment somewhere to go

Bencic’s product writing shows the same expanding circle. In April 2023, he described a move beyond standard blockchain nodes, the software through which applications communicate with networks. Tenderly was combining network virtualization and node customization so teams could work against production data in environments where they had more control. The aim was to remove infrastructure chores from the developer’s desk.

In November, he invited developers into early access for Virtual TestNets. The invitation treated testing infrastructure as something users should help shape. Separate frontend, backend and smart-contract work needed a place to meet. A shared environment could expose the disagreements between those pieces before customers had the pleasure of discovering them.

The April 2024 launch made the argument concrete. Public test networks could leave application teams wrestling with scarce test tokens, incomplete data and repeated migrations. Virtual TestNets offered customizable environments synchronized with mainnet state, with built-in inspection and debugging. The proposed improvement was ordinary in the most useful sense: spend more time testing your application and less time arranging the test.

For Bencic, that development extended an old concern rather than introducing a new personality. TxFlow made a transaction visible. A virtual network made an application’s surroundings available for experiments. The scope had grown from one execution to a working environment, but the intended benefit remained another informed attempt before the expensive one.

The morning after launch

In June 2026, the emphasis widened again. Bencic announced Tenderly’s positioning as a simulation company for onchain operations. Engineering remained in the picture, joined more explicitly by risk, finance, support and operations teams. The platform was being redesigned as Console, while its simulation tools were brought together to help teams model activity under production conditions.

The following day, his writing turned to what happens after a product launches. A smart contract can be carefully checked and still operate among dependencies its authors do not control. Prices change. Other protocols change. People update parameters. Bencic organized the problem into four kinds of risk: code, market, process and dependency. A launch date does little to retire any of them.

His public reflections on Consensus 2026 added a useful admission. He had expected conversations with traditional finance to feel like two industries talking past each other. Instead, he came away wanting blockchain teams to borrow more of the institutions’ attention to practical operation, responsibility and adoption. An engineer interested in removing friction was finding value in people paid to ask awkward questions.

That is a fitting place to find Bencic eight years after Tenderly’s founding: still arguing for a chance to look before acting, with a broader definition of who needs the view. The hackathon team once wanted a better way to finish its project. Its CEO now wants organizations to keep rehearsing long after theirs have launched. Tenderly kept the name of its first idea. Bencic kept the habit of giving the next idea a test run.