There is a type of corporate memo that makes a great effort to sound painful but ends up sounding triumphant. One of those was shared in early May by Jarek Kutylowski, CEO and founder of DeepL, a company based in Cologne. More than a fifth of the company’s workforce, 250 employees, were being let go. The explanation was succinct, straightforward, and full of language that has become practically ritualistic in Silicon Valley and its European neighbors: smaller teams, fewer layers, AI embedded at the center of everything. mode of the founder. The phrase appeared in the memo as if it belonged there.
Although DeepL is not as well-known as Google Translate, it has long had a significant influence in business technology circles. Established in 2017 as a more sophisticated and nuanced substitute for the pre-existing translation tools, it gained a reputation for managing linguistic nuances that other systems struggled with—the kind of thing a native speaker would notice even when a machine didn’t. The company raised $300 million and reached a $2 billion valuation by 2024, while IPO rumors lingered in the background. That appeared to be the course of a growing business. Because of this, the May announcement seems, at the very least, sudden.
Kutylowski presented it in a different way. In his LinkedIn post, he characterized the layoffs as a purposeful early move rather than a retreat. He suggested that most businesses acknowledge the need for this kind of decision but lack the courage to carry it out. He seems to really think this. The memo doesn’t have a defensive tone. It reads like a manifesto from someone who, instead of responding to a wave, has persuaded himself that he is ahead of it. That might be the case. It’s also possible that the wave has already arrived and the response has simply been dressed up in strategic terminology.
The figures themselves are striking. The organization’s headcount decreases from slightly over a thousand to about 750, and the reductions are not limited to a single function. A flatter organizational structure with fewer levels of management, quicker decision-making, and increased autonomy for smaller teams is the declared objective. DeepL is not the only company with this vision; Mark Zuckerberg of Meta proposed having a single employee perform the tasks of entire teams. Around the same time, Coinbase announced its own cuts, including clear limitations on management depth. The fact that the playbook is consistent across time zones and business sizes is either proof that it is accurate or proof that it spreads easily.
It’s worthwhile to consider DeepL’s current competitors. Google Translate was the company’s primary competitor at launch, and DeepL outperformed it in terms of quality. In 2026, the competitive landscape is more chaotic. ChatGPT is a translator. Claude does the translation. Some users claim that DeepSeek does a great job with German. People are already paying for and utilizing general-purpose AI systems that have subtly incorporated the services that previously required a specialized translation tool. Regardless of the wording used in the memo, that is a real and likely uncomfortable shift.
As this develops, the more difficult question is not whether DeepL made the right decision regarding headcount, but rather what the company is now and what it will become after the restructuring is complete. The AI agent tools and real-time speech translation that Kutylowski mentioned as the company’s next phase are truly intriguing. Given how quickly all of the major players are moving, it is still unclear if DeepL will be able to establish a sustainable position in those markets. The memo exuded confidence. Memos tend to be more courteous than the market.
250 employees were informed that their work was no longer organized in the same manner. There is no need for a buzzword in that section. It just needs to be acknowledged that there are real people in Cologne clearing out desks behind every reorganization that is announced on LinkedIn.