TSMC Interview Questions
& Process

The job post and a 1-hour interview with the hiring manager, followed by a 2-hour team interview in San Jose, suggested they were looking for a full-stack AI engineer capable of building products, particularly in RAG and Agentic AI. The salary for this San Jose role was below average. The hiring manager in San Jose was humble, curious, and attentive. The interviewing team in San Jose seemed quite junior, with even their newly hired "Sr. AI Engineer" having only a year of industry experience in finance and apparently no AI production experience. The most negative experience was with their Taiwan team during a "tech interview." The recruiter's invite mentioned a 15-minute presentation followed by a 45-minute Q&A. Seeking to prepare effectively, especially for the Taiwan team, I asked the San Jose hiring manager for feedback. I prepared a 15-minute talk showcasing my production AI experience in RAG and Agentic AI, including some industry-first work, and anticipated potential discussions. My preparation was highly relevant to the job description. However, during the actual interview, the two interviewers did not allow me to present at all, even after I mentioned my prepared presentation within the first three minutes. They ignored my request and proceeded to ask questions loosely related to the job. I did my best to answer: 1) Regarding an ambiguous "real case" scenario, I detailed my real-world approach, but the discussion lacked depth and they didn't seem satisfied. 2) They then focused heavily on time-series questions related to my 1-year experience in supply chain from over 4 years ago, which I recalled and answered comprehensively. 3) They asked about my fraud detection experience from over 5 years ago, which I also answered to the best of my ability. All questions were traditional ML, with no connection to GenAI. The interviewer claimed he wanted to assess my communication, but in my view, he lacked significant ML production experience, possibly none hands-on. He seemed difficult to please, and the discussion didn't go deep, feeling more like he was probing his own limited knowledge than engaging in a curious or tech-savvy dialogue. He even asked conceptual questions, which is unusual for hands-on roles unless absolutely necessary. I questioned the relevance and depth of these questions for assessing communication or technical skills. Having interviewed many people in AIML roles, I found his questions shallow. I never got the chance to present as initially required. Throughout the hour, nothing related to GenAI, the very reason for my application, was discussed. When I inquired about RAG and agentic AI, I received a biased response. The interviewer, who has no finance experience (checked LinkedIn), strongly asserted that Agentic AI wouldn't benefit TSMC's BI decisions and claimed TSMC isn't rule-based, unlike banks. This is baffling as I have extensive AIML experience in banking and have worked on broader cases than in other tech companies. It's contradictory to hire for GenAI roles while expressing such skepticism. My presentation could have addressed his biases if he had allowed it and shown more curiosity. The entire experience highlighted strong bias, a lack of curiosity, and poor collaboration potential. I also asked about AI collaboration practices, but the interviewer's response was vague. I question how such a person could evaluate my communication, for which I have clear records on LinkedIn. I provided candid and polite feedback to the recruiter and hiring manager post-interview. Given this experience, I no longer wish to join, as the Taiwan team gave me a sense of potential toxicity in daily work. If there's a lack of belief in AI's potential, why hire for it, especially with a decision-maker expressing such disapproval? It hinders AI work, particularly product development. Despite needing this job as a Canadian needing to relocate to the SF Bay Area urgently, my efforts felt wasted on biased, disqualified individuals. I'd rather invest more time finding a place where I can enjoy the work and collaborate with a passionate team.

First, HR will call you to ask some basic questions about how well you know the semiconductor industry. They'll probably ask about semiconductor manufacturing steps and the company's past. During this first part, feel free to share your thoughts on TSMC's recent investments and industry trends. Second, assistant managers from the department you're potentially joining will set up a formal meeting. They won't dive deep into technical details, but they will want to see what you know and will mention that it's a demanding role. They might bring up work-life balance, and you'll really get a sense of it once you're working in the factory. The last step is with another HR specialist who will do a final background check to make sure all your application details are accurate. If you pass all these stages, you've likely secured the job and will probably get an offer.

So, I applied for a Power BI Developer position, interviewing in October 2025. The first round was pretty great, the interviewer was super energetic and asked some really good follow-up questions, so it felt very interactive and positive. I moved on to the second round, which was supposed to be an hour-long panel interview. But, honestly, this round was a bit of a letdown. The two interviewers seemed kind of checked out and not really ready, which made the conversation way less engaging than the first one. It started off strong, but the second round just didn't have the same vibe or structure as the first.

I had one video interview with a team member, then another with their boss. Neither seemed to have read my resume. They had me go over my resume and asked me about my past experiences. They really wanted to know if I understood the technical background and theory behind the projects I'd worked on. Both interviewers really stressed that this job would include overtime more than twice a week, and also some weekend and night shifts. They asked me a lot about how I felt about that and how loyal I'd be.