The first machine was a 286, beige and modest and now old enough to deserve a small plinth in a museum. Martin Klenk’s father was a computer scientist who worked practically with machines; his mother worked in the city library, where the shelves held programming books and reference manuals. Between them sat a boy with permission to experiment. He learned gradually and, as he remembers it, playfully. There was no immaculate origin scene, no flash of destiny. There was a computer, a book and the pleasant discovery that a system would answer if you asked it the right question.
By secondary school in Herrenberg, in southwest Germany, curiosity had become organization. Klenk led a Linux working group, helped build and operate a student internet café and organized Linux installation parties. The phrase belongs to a more cheerful age of computing, when persuading a machine to accept a free operating system felt sufficiently festive to require company. His headmaster let the students travel to trade fairs and Linux days. The school understood a useful thing early: technical ability becomes more interesting when it leaves the bedroom and acquires collaborators.
A wall full of evidence
Klenk arrived at the Technical University of Munich with practical fluency and a wish to give it scientific foundations. He studied computer science, specializing in databases, algorithms and the emerging machinery of big data. He also freelanced as an IT security trainer and joined Academy Consult München, a student consultancy that introduced him to people beyond what he called his own “bubble.” There he met Bastian Nominacher and Alexander Rinke, whose skills in information systems, finance and mathematics complemented his own.
His school record had already suggested that abstractions did not frighten him. In 2006 he finished with the best Abitur result in his year, received the Ferry Porsche Prize for mathematics and physics, and won first prize in Germany’s national mathematics competition. A scholarship supported him from 2007. He was not, however, being groomed in a laboratory jar. Between school and university he completed civilian service in Munich, and during his studies he kept one foot in paid technical work. Theory was something to be tested against systems that had users, deadlines and inconvenient habits.
In 2010, the three took on a project for Bayerischer Rundfunk. The broadcaster wanted to understand and improve an IT service-management process. Conventional analysis was not getting them far. Yet the system itself had been recording events all along: what happened, when it happened and in what sequence. Those logs were a procedural autobiography with terrible prose and excellent recall.
The students applied process-discovery algorithms inspired by the research of Wil van der Aalst. To show the result, they printed more than 300 process models on A3 paper, covered a conference-room wall and spent a day with the people who owned the process. One of the pleasant ironies of Celonis is that a company devoted to revealing digital operations made its formative argument with a monumental quantity of paper.
The clue Celonis followed
The prototype did something the standard methods had not: it allowed the customer to see the process as it ran, not as a tidy diagram claimed it ran. Klenk and his collaborators recognized the distance between academic promise and enterprise usefulness, then decided to cross it. On June 15, 2011, they formalized Celonis. They began in Nominacher’s apartment, three recent students with a product whose subject was the untidy interior life of large organizations.
Klenk had started a computer-science master’s degree in 2010. The company interrupted it. He left without completing the qualification, a choice that looks obvious only after the outcome is known. At the time, Celonis was not a celebrated category leader. It was a very small company trying to persuade large institutions that a technical research field could answer expensive operational questions. The risk, as Klenk later put it, was manageable because the founders had few obligations. Manageable is not the same as imaginary.
“With my heart, I am an engineer, and that is what I really enjoy: building software, developing products, solving problems.”Martin Klenk
The expensive gift of losing
Early success did not arrive in a neat line. The young company lost an important prospective customer. Rinke was crushed. He asked Klenk to go with him for two weeks to the Baltic coast, away from the ordinary noise of the business. Their response to rejection was not a prettier pitch deck. They rebuilt the product. It had been too technical. It lacked the moment in which a user could see, quickly and almost physically, why the software mattered.
Rinke later said that without that rebuild Celonis might not have survived. The anecdote reveals the practical temperament of the technical founder. A product is not a monument to the intelligence of its makers. It is an agreement with its users, and users are under no obligation to admire the difficulty of the code. Klenk’s public descriptions of Celonis return repeatedly to usefulness, accessibility and the needs of customers. Elegant machinery hidden behind an obstinate interface is merely a private hobby with invoices.
The company found demanding allies. Siemens became an early customer, testing software built by young founders against the disorder and scale of a major industrial enterprise. That pressure was productive. If the architecture could begin at a high standard, it could later be simplified; a small design pushed upward often breaks. Customers also financed the company. Celonis went roughly five years without outside investment, an unusual sequence in software: first a useful product, then revenue, then professional investors.
As CTO, Klenk led the technical development while Nominacher and Rinke carried complementary responsibilities. The founders adopted a rule for decisions: once made, a choice belonged to all three. They would not reopen a mistake merely to establish who had originally voted for it. It is a mature policy, and also a superb device for making meetings less operatic.
Leads a school Linux group and helps operate a student internet café.
Builds the Bayerischer Rundfunk prototype with Nominacher and Rinke.
Co-founds Celonis and takes responsibility for technical development.
Shares Germany’s Deutscher Zukunftspreis with his co-founders.
Speaks about data products, data lineage and operational context for AI.
A prize for making theory behave
Process mining sounds obscure until one considers the daily absurdities of an organization: an invoice waiting in the wrong queue, a shipment circling a manual exception, a purchase duplicated because two systems disagree. The software uses event data to reconstruct these paths, compare the intended process with the real one and expose bottlenecks or deviations. The founders often compared it to an X-ray. The patient, in this case, is an accounts-payable department with seventeen passwords.
Klenk helped push conformance checking from a research concept working on small data sets toward enterprise use. He also co-authored work on Celonis Process Query Language, built to express business questions about process data. By 2019, the work had reached its fifth broad product generation. That year, Germany’s president presented Klenk, Nominacher and Rinke with the Deutscher Zukunftspreis for turning process mining into a practical technology for work and value creation.
The award belonged to a longer technical migration: from a narrow event log to richer transactional and master data, from a map of one process to an end-to-end view, from diagnosis toward recommendations and action. Celonis became a company with offices across countries and customers across industries. Its private valuation rose sharply, bringing the inevitable arithmetic of founder stakes and billionaire lists. Klenk’s more revealing sentence remained the older one. At heart, he said, he was an engineer.
“Good products come and go, good teams prevail.”Martin Klenk
The team after the product
By 2023, when Klenk spoke at Slush about taking a European company global, the problem was no longer whether the software could work. It was how a technical organization could keep working as it spread across continents. His answer has been notably social for a man known for algorithms. People, he says, want to be good at what they do. The company’s task is to create an environment where an engineer can be the best engineer possible, a product manager the best product manager, and colleagues in sales or marketing equally capable of their strongest work.
This is not decorative culture talk. It follows directly from his view of products. A current release will age. A strong team retains the ability to notice, rebuild and make the next one. Klenk has said Celonis can keep its pioneering position only through continuous development. Stand still, he warns, and being caught becomes a question of time.
His recent subjects widen the old idea without abandoning it. At an international process-mining conference, he discussed applying the process lens to unconventional problems such as data lineage. At the Passau Data Summit, he addressed data products. With AWS in Hamburg, he explored the operational context that artificial-intelligence systems need if they are to act usefully inside a business. The vocabulary evolves. The habit is familiar: read the traces, connect the events, make the machinery comprehensible.
Outside work, a 2019 interview found him riding with a regular road-cycling group before the office or into the countryside around Munich. The detail suits him almost too neatly. A bicycle rewards balance, repeated motion and small mechanical adjustments; hills are unimpressed by valuation. More importantly, it is done with other people.
The career that began beside a 286 has acquired scale, prizes and a considerable corporate apparatus. Its governing pleasure seems largely unchanged. Klenk likes to puzzle things out with others. The machines keep diaries. The engineer reads them. Then the team decides what to build next.