Process
I applied through a staffing agency.
The process was generally smooth and professional in terms of communication. However, there was a significant mismatch between how the technical interview was described and what was actually evaluated.
Ahead of the interview, I was explicitly told multiple times that the round would be heavily focused on theoretical and mathematical aspects of reinforcement learning, including a whiteboard-style session. Based on this, I invested a substantial amount of time preparing in depth, approaching it almost like an academic-style examination.
In practice, while some theoretical questions were asked, there was no whiteboard component, and a large portion of the discussion focused instead on practical, open-ended applications of reinforcement learning in real-world scenarios in the context of LLMs. This misalignment made my preparation largely inefficient and left me feeling that I had focused on the wrong areas. Given the level of preparation required, this was quite frustrating and gave the impression of being set up for a different evaluation than the one that actually took place.
Was this helpful?
Interview Statistics
The following metrics were computed from 1 interview experience for the DeepL Machine Learning Scientist role.
Success Rate
DeepL's interview process for their Machine Learning Scientist roles is extremely selective, failing the vast majority of engineers.
Experience Rating
Candidates reported having very negative feelings for DeepL's Machine Learning Scientist interview process.
DeepL Work Experiences
- 1.0Brilliant colleagues overshadowed by attrocious leadershipSenior Software Engineer • Current EmployeeMay 12, 2026Pros: - Brilliant and dedicated colleagues - An interesting area of workCons: - Leadership is an absolute abomination, has no empathy, and becomes increasingly detached from the workforce every time they open their mouths. - There is a lack of transparency on decision making, and the l-team is straight up lying about things. - There is no accountability for l-team bs. - There is a lack of career options, and goals and criteria for promotion are ever changing. - There is a reorg every 10 minutes, with line managers changing every other week.
- 5.0Learning opportunities with travel involvedQA Engineer • Former EmployeeWrocławMay 7, 2026Pros: A lot of opportunities to learn from the best. A product used by millions worldwide.Cons: A lot of travel. But it could be in the 'pros' category if one likes to travel.
- 1.0Leadership need to wake up before it's too lateSenior Software Engineer • Former EmployeeMarch 28, 2026Pros: The product continues to be excellent, despite everything.Cons: The company has gone steadily downhill since the (no longer) new Chief People Officer was appointed. Layoffs disguised as performance issues. Engagement scores steadily plummet, with nobody listening. More and more colleagues taking time off sick with stress. Attrition is alarming—Jarek says it's 'industry standard' but he should look more closely at who is leaving: • Principal Engineers • Senior Engineering Managers • ERG Leads • The Culture Programs Lead • Staff Engineers • People who have been champions of DeepL's culture and mission for a very long time and now no longer feel they can work there This is a sign that something is deeply wrong—as is the fact that the German colleagues now feel they need a Betriebsrat to prevent further mistreatment by leadership.
