I went through a multi-stage interview process for a senior engineering role. The early stages were well-organized — communication was clear, and initial conversations were professional and aligned with the company’s stated values. The take-home assignment required a significant time investment and included real-world challenges (data processing, performance considerations, deployment). However, there were issues with the provided infrastructure (broken Kubernetes credentials), which required additional effort to work around. This part of the effort was not acknowledged during the process. The final interview focused heavily on a single implementation detail. After identifying a minor issue, I acknowledged it and discussed how it could be addressed. Despite that, the conversation remained focused on the same point for an extended period, rather than moving on to broader topics such as system design or architectural decisions. The feedback I later received stated that I had difficulty accepting criticism and perceived questions as personal. This was surprising, as during the interview I explicitly acknowledged gaps and engaged with the feedback in real time. Since the interview was recorded with an AI note-taking tool, I requested the transcript. After reviewing it, I did not find evidence supporting the claim that I was defensive or resistant to feedback. Instead, the transcript shows that the issue had been acknowledged and discussed, yet the conversation continued to revisit it repeatedly. This created a disconnect between the documented interaction and the final feedback.

I completed the HR screening interview, which went smoothly. The technical interview, however, felt unusual. The interviewer asked questions about PostgreSQL locking mechanisms, the Kubernetes ecosystem, and microservices architecture. While I answered concisely, there were almost no follow-up questions, which made it difficult to understand the expected level of detail. During the conversation, I noticed that the interviewer seemed to rely heavily on an AI assistant while asking questions. Because of this, it was challenging to have a natural technical discussion, as follow-up questions or clarifications were rarely provided. This made the interview feel less like an evaluation of real-world experience and more like a checklist-based process. Afterward, I received feedback stating “weak understanding of PostgreSQL locks,” which was surprising given that I maintain open-source projects that depend extensively on PostgreSQL locking and transaction isolation levels. With clearer communication or more interactive discussion, I believe the interview would have reflected my actual knowledge much better. Overall, the process left me with mixed impressions, mainly due to the lack of technical engagement during the interview.