1p3a Experience · Oct 2025 · San Jose

TikTok Machine Learning Engineer Interview Experience and Outcome

MLE coding Easy
1 upvote 2 replies

Interview Experience

I need some points! I want to accumulate more to read interview experiences. This happened in early August when a TikTok LinkedIn HR contacted me. I was interviewing for another TikTok Researcher grad

Full Details

I need some points! I want to accumulate more to read interview experiences. This happened in early August when a TikTok LinkedIn HR contacted me. I was interviewing for another TikTok Researcher graduate position at the time. After I sent her my resume, she told me that TikTok only allows one interview at a time, and she could only move me to the MLE position if I stopped interviewing for the other position. I didn't reply to her then. Two days later, my Researcher graduate application was

rejected. I quickly replied to her on LinkedIn that I was now available again. Two days later, she told me my resume had been approved and I would be moving on to the interview stage. Then we scheduled an HR call a week later. This HR person was probably in China, as we communicated entirely in Chinese, discussing my resume. I had always thought HR calls weren't for screening applicants, and I hadn't prepared, so my Chinese answers were very poor and hesitant. I guess I haven't practiced interviewing in Chinese; I'm probably better suited to English... Then an interview was scheduled for mid-September, which was really late; I don't know why the interview scheduling took so long. The interviewer was a Chinese person working in the San Jose office, and the entire interview was conducted in Chinese. The interview process consisted of a 30-minute ML interview and a 30-minute coding interview. The ML interview asked some project-related questions. It seems like TikTok is currently looking to develop agent-related technologies internally, as both interviews involved a lot of agent-related knowledge, but many were open-ended questions. I tried my best to answer, but my abilities are limited; I could only offer some fine-tuning RAGs and basic transformer answers, along with the project-related questions from my resume. The coding interview included an ML coding question, which was the first time I'd encountered ML coding in an interview. Since I'm an SDE looking to switch to MLE, I didn't expect the coding section to include ML coding. It was roughly about achieving something using gradient descent, which I didn't understand at all. I said I could use binary... Let's do the search. He said this wasn't gradient descent, and then it dawned on me. I finally managed to figure it out, albeit with some stumbling and difficulty. Actually, I felt the interviewer was quite friendly overall, much better than the cold, impersonal interviewer from before. I also spoke more fluently and thoughtfully about ML-related knowledge, but maybe my abilities are still limited. Finally, I was

rejected a week later. It's always the first round of rejections... I'm quite heartbroken because I felt I had given it my all in the interviews, and it was better than the last one. I thought I could at least get a second round... Later, I discovered I seemed to have been blocked by TikTok. Then, an HR person on LinkedIn reached out to me for a lower-bar position that was a better match for me overall, but they stopped interviewing me...

About This Question

This is a candidate experience report from a bytedance interview for a mle role during the coding round reported in 2025.

Difficulty rating: Easy