Daniel Dines on Building UiPath, Risk and Finding Peace

Harry Stebbings asks Daniel Dines what he was like as a child. Dines describes himself as curious and argumentative, sometimes too argumentative for his parents. He enjoyed debate and having an opinion, although over time he has come to prefer listening. Curiosity, he says, has stayed with him.

He did well in primary school but later became an average student, distracted by friends and other activities, with one exception: mathematics came naturally. He chose mathematics and computer science at university in Romania in the early 1990s, but found an atmosphere that still reminded him of the communist period. After the first week he decided to attend mainly for exams.

Dines became passionate about bridge. The best books he could find were in English, which he did not yet speak, so he and a friend pooled their scarce money for an American bridge book. He read it with a dictionary. Years later he worked with a dialect teacher to improve the pronunciation that he had largely learned from the written page.

His family could no longer support him when he was 19, so he took on assorted work, including buying dollars in Bucharest and selling them elsewhere for a margin. He also started a small business collecting newspaper job advertisements, calling employers and matching students with work. It supported him for about a year, earning roughly $1,000 by his recollection.

Before his first programming job he lived on about $30 a month and sometimes went without food. Dines says that experience left a lingering anxiety around money. It could keep him grounded, he thinks, but it could also hold him back and prevent him from moving as quickly as he might.

After almost ten years bootstrapping UiPath, raising its first outside capital brought both responsibility and freedom, Dines says. Other people had trusted him, and the money loosened the internal brakes he associated with having to do everything himself. For a period afterward he felt almost fearless about pursuing a much larger opportunity.

He recalls negotiating a bridge round while the company was close to running out of cash and already had about 100 employees. The investors ultimately agreed to an uncapped instrument with a 25% discount. Looking back, Dines says he did not seriously contemplate the possibility that negotiations might fail and the company could run into trouble.

Stebbings asks whether surviving on a dollar a day made him willing to take big risks. Dines thinks knowing he had already lived through hardship may have helped: once the opportunity appeared, he felt he owed it to himself to pursue something big, rather than settle for what had seemed attainable before.

Dines learned to code after a friend told him a programming job paid $300 a month, ten times what he was living on. A girlfriend gave him a photocopied C++ book by the language's creator. Without a computer, he read it like an adventure story and worked through the concepts in his head.

He says his strength was reasoning about code and debugging, more than typing out a solution once he could already see it. A theoretical foundation, reinforced by mathematics, gave him a way to think about whole systems. He still reads about programming languages and connects ideas from different systems to problems his own team faces.

Dines finds it harder to explain the systems he sees in his head: his thoughts move faster than his speech. Working with UiPath's investor-relations team taught him a useful structure for public-company communication: state the main points at a high level, then expand each one and offer to go deeper where the listener wants detail.

His first programming employer hired him after a conversation rather than a coding test. The small firm had no spare computer during the day, so he worked at night. That gave him access to its most powerful machine and a quiet room, as much as he wanted. For six months he often worked 12 to 14 hours a day and saw few friends.

Dines remembers the period as one of the happiest of his life. Programming gave him the feeling of creating something and entering a state of flow; concentrating on a computer for half an hour could quiet his anxiety. He misses that depth of focus, although he still enjoys thinking about product decisions and occasionally writing poetry.

He believes continuing to change is evidence of learning. At the time of the interview he felt he had learned more in the preceding 12 to 18 months than in the previous five years. He contrasts that with his early thirties at Microsoft, when he felt stuck and unhappy after leaving his comfortable life in Romania for the United States.

Dines had pushed himself to take the Microsoft job in pursuit of financial freedom, but felt he was waiting out a prison sentence. In retrospect he thinks he could have accepted the decision and enjoyed the opportunities it offered, from learning and meeting people to improving at sport. His lesson is that the preferable choice can also become a pleasurable one.

He eventually left Microsoft and returned to Romania with about $150,000 in savings after five years. It was enough to live for a while without doing work he disliked, but building a company seemed the way to make the move sustainable. Friends had not believed he would leave and return with relatively little money.

Stebbings says the amount of money that seems like freedom can keep rising as one's circumstances change. Dines admits that even he retains some anxiety about money. He suggests that finding people and an environment that recognize your professional and human value can make it easier to trust that you can earn and enjoy the work itself.

Dines says he both enjoys his work and feels its burden. As an empathetic leader, he hears negative reports and internal problems that he cannot always act on. The CEO's position can be lonely: even trusted friends and colleagues do not experience the responsibility in quite the same way.

He attributes some of his biggest CEO mistakes to impostor syndrome. Doubting himself, he tended to hire executives for their experience and sacrifice personal chemistry. With rare exceptions, he says, that trade-off went badly. He has found the popular advice to fire quickly harder in practice than it sounds.

Senior executives need freedom and time to build their teams, so by the time signs of a mismatch are clear, they may already have created layers beneath them. Dines says giving negative feedback is difficult for him, especially when he does not feel a strong personal bond with the recipient. He is still working on that skill.

Investor Carl Eschenbach told him that firing people should be hard: losing that discomfort, in his view, would make someone a worse CEO. Dines agrees. He also accepts criticism that he has sometimes given leaders freedom without inspecting their work closely enough; inspection, he argues, need not remove autonomy.

In product and engineering, Dines probes decisions with repeated questions: why was this chosen, and does the original reasoning still hold? Going deep can reveal a problem early. He feels more fluent in those areas than in go-to-market, finance, HR or marketing, which he describes as languages he has had to learn.

Asked about executive turnover, Dines says people have ranges in which they can operate effectively, and a hypergrowth company may outgrow those ranges. Some people leave when a more experienced leader is hired above them; others stay, learn and earn a larger role. He estimates that UiPath went through four or five executive teams before going public.

UiPath bootstrapped for about a decade. Dines says he initially did not believe investors would fund a company with so little to show, and in the mid-2000s venture capital felt distant from Romania. He does not regret the time: those were formative years in which he learned to navigate adversity and build a business.

Doing nearly everything himself taught him to write concise Google advertisements, create software licenses, sell, support customers and handle practical operations. He worries that abundant funding can let founders delegate those lessons too early. Stebbings recognizes the pattern in 20VC, which he says recorded 450 shows before earning money.

Dines says UiPath raised its first money in 2015 and ended that year with roughly $500,000 in revenue. He dates its robotic process automation product-market fit to 2014. Before that, the team had built a low-code tool that could automate clicks and typing through a computer interface, but had not understood the size of the market it could serve.

The first version of UiPath Studio launched in March 2013. The company's earlier customers were developers buying an SDK, and they showed little interest in the lower-code product. Then a middle manager at an Indian business-process outsourcing company found it while looking for an alternative to Blue Prism, whose approach he disliked.

The inquiry arrived from a personal Yahoo address and almost went unanswered because UiPath did not normally work with individuals. Dines nevertheless asked a developer to give a demonstration. She discovered that the caller worked for a large company, and Dines joined the conversation. Asked how UiPath compared with Blue Prism, he looked it up and frankly said his startup was much smaller.

The prospect invited UiPath to India to demonstrate automation at cost. Dines sensed an important opportunity and sent three people, including his then CTO, to Chennai for about three months—a large commitment for a team of roughly seven. They combined UiPath's software with a Microsoft orchestration product to deploy and monitor the bots.

The Indian company became a customer and, Dines recalls, generated about $100,000 in revenue in 2014. More consequentially, the project showed him how much repetitive work remained inside very large companies despite existing business software. UiPath could address that need if it built the orchestration around its automation tool.

Stebbings points out that this discovery came roughly nine years into the company's life, long after founders are usually told to identify a customer and problem. Dines agrees it was unusual and says he would not recommend relying on the same route. His team built the software because it loved the engineering problem, then found a large use case.

Dines thinks the market might eventually have found UiPath without the Indian inquiry, but perhaps one or two years later, with a different outcome for its size. He recalls the 2015 seed round as $1.6 million at about an $8.1 million post-money valuation, which he considered somewhat cheap given its revenue, while recognizing that the company was in Romania.

He met investor Dan Lupu in 2014 and was initially skeptical about outside money and the proposed terms. Negotiations stretched from a term sheet around August 2014 until money arrived in July 2015. By then, Dines says, interest from large prospective customers had strengthened UiPath's position, and the investors bought common stock with terms tied to the next round.

Dines remembers being impressed by Accel's London offices when UiPath later sought capital there. He wanted to associate the company with success. The discussion brings back a contrasting memory from Microsoft in the early 2000s: an email announcing the removal of towels from some employee showers as a cost-saving measure.

To Dines, the towel announcement signaled that management had underestimated the message such a small cut could send to employees. He says he now thinks carefully about the meaning of communications, and tries to speak authentically because pretending or obscuring reality does not come naturally to him.

When announcing that Rob Enslin would become sole CEO, Dines told leaders that clarity is a central leadership responsibility. He used the biblical story of Adam and the forbidden tree to argue that stating a rule is not enough: leaders should also explain the reasoning behind it and identify the territory where the rules are still uncertain.

The colorful analogy did not necessarily land as he intended, which taught him to choose metaphors his audience can absorb. But he stands by the principle: explain decisions and distinguish known boundaries from uncharted territory, so people know where they can rely on a compass and where they must explore together.

In 2019 UiPath had substantial cash and Dines planned to burn at most $150 million while investing aggressively for growth. His rough plan aimed to build capacity for $600 million in revenue, accepting that a lower outcome could still justify the investment. Then a finance leader warned him that the burn could reach $300 million.

The news was frightening because, Dines says, overspending could put the company at risk. At an emergency meeting in August, leadership decided to stop hiring and reduce staff. He recalls laying off around 400 people out of roughly 3,000, although he is approximate about the percentage. The decision was especially painful when the broader economy was still strong.

A competitor celebrated UiPath's difficulties, Dines remembers, but continued hiring into the pandemic. In his account, the cuts left UiPath leaner as COVID arrived. The lesson he draws is that a leader must sometimes accept pain and reputational cost to do what they believe is necessary for the company; he does not minimize the difficulty of telling employees.

Asked about risk, Dines says his response varies. Sometimes he is fearful, while at other times he is drawn to the upside and makes large bets without dwelling on consequences. He thinks the switch toward taking bigger risks began with the first financing round, before he had accumulated his later wealth.

Dines initially declined Sequoia's interest in UiPath's Series B because he did not feel chemistry with Carl Eschenbach after their first meeting. Rather than hide the reason, he told him directly. Eschenbach became curious, they kept talking, and the candid exchange became the foundation of a relationship Dines came to trust.

Dines says Eschenbach later made clear that if Sequoia invested in the RPA market it would choose UiPath. He also acknowledges the signaling effect of an investment from the firm: some people follow Sequoia and look specifically to join its portfolio companies.

Dines says he wanted stronger go-to-market leadership and saw Rob Enslin as an exceptional candidate; giving him the CEO title was the way to bring him in. He did not see the title change as a threat to his ego, having already built the company from nothing to more than a billion in scale.

He had been open to replacing himself much earlier. During the first fundraising discussions, an investor introduced a potential CEO because Dines seemed more like a technical founder. Dines introduced the candidate to the team, but their joint pitch did not work and the investor asked him to pitch alone again.

Public speaking had once frightened him. Dines says the turning point was reframing a presentation as a duty to the company and its people rather than a test of his own ego. That let him direct his nervous energy toward delivering a message; he may start slowly, but finds his rhythm as he reads the room.

Stebbings asks how to detach happiness from a company's good and bad quarters. Dines says he does not think of himself as a consistently happy person and seeks peace of mind and impact instead. He feels most at peace when he can focus, as he did while programming, and finds some calm in meditation or prayer.

On work-life balance, Dines offers his own strongly held view: building a company occupies his thoughts even when he is away from work, so he finds strict boundaries difficult. He describes his mind as obsessive and says he sometimes has to tell himself to stop when thinking has become unproductive circular rumination.

Stebbings worries that powerful incumbents will leave little room for AI startups. Dines shares some of the concern and thinks the early application market may be difficult for smaller companies, while acknowledging that friends see opportunity in vertical AI. He presents this as a judgment about an evolving market, not a settled outcome.

Dines says interest in AI can help UiPath by revealing more automation opportunities and making end-to-end workflows possible. It also creates competing narratives that AI will replace automation entirely. He is not persuaded that current technology has exhausted its potential; in his view enterprise applications are still at an early stage.

At the time of the interview, he saw coding autocomplete as a prominent use case and many other enterprise projects still in proof-of-concept stages. He expected document reading and summarization to be useful, especially paired with automation. UiPath, whose business is automation, has an interest in that outcome.

His most memorable moment as CEO was establishing UiPath in Japan, where growth for a period exceeded that in the United States. In 2016 a Japanese executive contacted the company about bringing RPA to the country. When a bank opportunity arose, Dines changed his travel plans to meet the potential customer in person and build trust.

At SMBC he met a leader he recalls as Yamamoto-san and felt an immediate personal connection. Over a shabu-shabu meal, Dines promised that if the bank gave UiPath a chance, the company would invest in Japan and return with a local team. By April, he says, six people were working there, more than he had promised.

On a later visit he saw dozens of people at the bank and its consultants working with UiPath software. He remembers the scene with pride. Stebbings encourages him to pause and recognize those moments, something Dines says he does not often do.

In the quick-fire round, Dines says he has learned more about what triggers his anxiety and helps his attention. His worst decisions at UiPath, he believes, came from not trusting his instincts. Having grown up under a dictator, he is also concerned about authoritarian regimes where people in command cannot be challenged.

Dines has revised his view of himself as an operator. An operator, in his definition, is not merely a project manager tracking tasks, but someone who understands a domain deeply enough to probe and challenge decisions. On that definition, he now thinks he is good at operating.

He challenges the assumption that relevant experience alone makes a good hire. He also qualifies the startup maxim that speed is always paramount: scaling a half-built product with a large go-to-market team before product-market fit can destroy a company. He agrees with Stebbings that moving quickly to test and learn is different from scaling prematurely.

Asked whom he would add to the board, Dines chooses Paul Graham, whose writing he says shaped his thinking during his formative years. He does not plan his personal life ten years ahead. Given how much the world may change, he hopes simply to remain healthy, keep his mind sound and do good.