Nicolas Orban and the art of changing the plan

Nicolas Orban once had a fairly concrete assignment: help build Deliveroo’s business in Belgium. Restaurants had to join a platform. Companies had to become customers. Years later, at Conduktor, the subject matter would involve streams of data rather than lunch, but the commercial question would remain familiar. Who has a problem pressing enough to do something about it?

That question gives his career a useful thread. Orban is the CEO and co-founder of Conduktor, the software company he started in 2020 with Stéphane Derosiaux and Stéphane Maarek. Its first appeal was to developers navigating Apache Kafka. Its subsequent work reached the people responsible for keeping an organization’s data usable, controlled and available. Somewhere between those audiences, a founder has to decide how much of the original plan to keep.

A sales apprenticeship, with homework

Before Conduktor came Deliveroo and Meero. At the former, Orban worked in business development and corporate business; at the latter, he became Head of International Sales. His professional background sits in the commercial machinery of startups: accounts, operations, expansion. There is a certain education available in asking people to buy something. The syllabus tends to arrive without warning.

His public university record adds an intriguing earlier subject. He attended Université catholique de Louvain from 2013 to 2015, listing a thesis about drones and their implications for privacy and data protection. His options included international commerce and intellectual property. Data, rules and markets were already sharing the page, though a thesis title should hardly be mistaken for a business prophecy.

A recommendation from his former Deliveroo manager, Laurent Stevenart, describes a young employee asking for responsibility early. After six months of professional experience, Orban was tasked with launching the corporate business in Belgium. It is a small but specific glimpse of his working life before he had a founder’s title: a new commercial operation, with responsibility attached.

Three founders, one awkward technology

Orban’s desire to start a company preceded the company itself. He has described choosing fast-growing startups as places to learn, and meeting Derosiaux and Maarek through a mutual friend in Paris. The two Stéphanes knew Kafka’s difficulties directly. Their eventual partnership brought commercial experience into contact with technical knowledge and an existing developer community.

“I’ve always known I’d start my own company.”

Nicolas Orban · 2025

Kafka carries events between software systems as they happen. That makes it useful wherever waiting for a batch of information would be inconvenient: a shipment notification, for example, or a transaction needing attention. Working with those streams requires expertise. Conduktor’s early desktop application gave developers a graphical way to inspect and manage their Kafka environment.

The founding team also understood distribution. By 2021, Maarek had taught more than a million students through his technical courses. Accel’s November account of Conduktor recorded 60,000 users across more than 9,000 companies, acquired organically since the summer 2020 launch. Those figures describe adoption, rather than a ledger of paid contracts. They nevertheless establish that the founding problem had plenty of company.

The customer in the next office

A developer can love a tool while the team responsible for the wider system still has unfinished business. Conduktor heard from those central teams. Developers had gained a more convenient way to work; the people overseeing security, architecture and shared infrastructure needed help of their own. That feedback widened the company’s audience and changed the buying conversation.

More stakeholders bring more questions. A person inspecting a stream wants to see what is happening. A central team must also consider who should have access, what standards apply and how those standards survive the next application. The developer’s immediate task and the organization’s continuing responsibilities occupy the same system. Software has to accommodate both.

Orban’s account of the transition acknowledges longer sales cycles. That is a revealing admission from someone whose earlier career was spent in sales. Winning affection from a user and winning agreement across a company involve different work. The distance between them can contain several meetings, another budget owner and a calendar that has suddenly become terribly busy.

The cost of keeping every idea

The product record contains edits. Console arrived in October 2022, during a company offsite in Milan. Gateway followed in 2023. Alongside those releases, Orban’s retrospective lists offerings the company discontinued. Cloud had encountered customers who wanted an on-premise deployment. Testing had failed to gain sufficient market traction.

“We also failed,” he wrote. Three words that make room for an actual decision. An offering can absorb considerable effort and still fail to earn its place in the next year’s priorities. Stopping leaves capacity for something else; it also means accepting that previous work will not develop as intended. The retrospective gives the abandoned products names.

The company’s finances demanded a revision, too. Orban has said the easy-money atmosphere of 2021 influenced him, and that Conduktor let go of nearly half its team in 2022. He later described working more closely with finance and paying greater attention to efficient growth. These were consequential decisions affecting employees. A subsequent funding round does not make them a footnote.

Even the smaller product measurements carry the same preference for subtraction. In its 2023 review, the company reported cutting onboarding from five screens to one. A roadmap usually has plenty of volunteers willing to add the next feature. Four vanished screens are a quieter entry in the record, and a useful one for the person trying to get started.

Three nights will do

The same editorial problem appears in a less technical setting: the company offsite. For its fourth gathering, held in Cély outside Paris in 2023, Conduktor shortened the previous full-week format to three nights. Employees could book travel within guidelines and a budget. The schedule included an escape room and a cooking challenge alongside modified Olympic games.

Earlier events had leaned heavily on sports. Orban acknowledged that this did not suit everyone. Teams were mixed across countries, departments and seniority. Then French transport strikes required backup arrangements. The event went ahead, after changes to travel plans and meeting points. There is something reassuringly practical about a founder’s account that has room for railway disruption.

Another team story offers a different pace. Head of Design Claudiu Danaila recalled signing his contract and receiving an invitation the following day to meet the team in Bordeaux. He had only hours to pack and reach the airport. Joining a software company could involve an unexpectedly urgent relationship with a suitcase.

Danaila also described the effort involved in persuading colleagues that design mattered to business, beyond appearance. It is a useful companion detail to the founder’s commercial background. A company selling easier ways to work with a complex system still has to argue about what makes work easier. Good intentions do not settle the interface.

An ocean is a hiring problem, too

For Orban, New York offered access to the U.S. market while preserving a connection with Europe. He has explained that choosing the city helped maintain that link, culturally and logistically. The decision places him near customers without requiring the company to abandon the international character of its team. His public professional profile now lists New York.

Recruiting supplied less comfortable lessons. He has spoken about early hiring mistakes and the difficulty of interpreting American interview confidence through European expectations. His account favors experienced local leadership and time spent together. Quentin Packard, whose background included Splunk and HashiCorp, became an important sales hire in that process.

There is a distinctly commercial humility in this part of the story. Someone accustomed to assessing prospects had to improve his assessment of candidates. A new market changes the meaning of familiar signals. Confidence needs context. So does a résumé. Moving the founder closer to the work creates opportunities to learn those distinctions in person.

The next plan still needs a pencil

In November 2024, Conduktor announced a $30 million Series B led by RTP Global, with M12, Ansa and Accel participating. The stated uses included U.S. expansion, product development and customer success. Its announcement named customers including BMW Group, Flix, Capital Group and Lufthansa. These are organizations with substantial operations, where an infrastructure purchase has to survive scrutiny beyond the initial demonstration.

Orban’s stated ambition remains broader access to data streaming. His company-hosted biography puts accessibility at the center of his work. That ambition has to coexist with the requirements of enterprise buyers: security, responsibility and an understanding of who can do what. Making a system easier to use increases the importance of deciding how it should be used.

The appealing detail in his story is the willingness to record an amendment. A product stops. A buyer changes. An offsite loses a few nights. The founder crosses an ocean and discovers another thing to learn. Each choice makes the portrait more specific than the familiar photograph of an executive beside a funding number. Orban’s work continues in the decisions that follow the announcement, when the next conversation tests the plan again.