Envié una solicitud electrónica. Acudí a una entrevista en TikTok en feb 2025
Entrevista
The interview process is recruiter call, followed by 3 technical rounds (each with a leetcode problem, resume overview, and ML questions), and a final HR round. Each technical round lasts around one hour, and I received a mix of medium and hard leetcode questions. For the ML portion, I was asked to mathematically derive the cross entropy loss, and code a Transformer block.
Preguntas de entrevista [1]
Pregunta 1
Leetcode: Binary Tree Maximum Path Sum, Serialize and Deserialize Binary Tree, Search in Rotated Sorted Array
Started with a recruiter screen with simple checks. Then went through 3 rounds of interviews. First 2 rounds were live coding with deep dive on previous projects. 2nd round went really deep of every previous work I did from MLE perspective with being questioned of every possible decision choice. Last was interview with team manager.
Preguntas de entrevista [1]
Pregunta 1
Recommendation system questions. Coding problems were I remember leetcode medium level.
Surprisingly straightforward — I expected a tougher challenge for a machine learning role. After a quick recruiter screen, the first technical round focused on implementing K-means clustering, which felt familiar. Handling edge cases for empty clusters was tricky, though. What really helped me prep were the algorithm explanations on PracHub, which gave me confidence going in. The final interviews were a mix of problem-solving and behavioral questions, and in the end, I received an offer that I accepted. Overall, it was a decent experience.
Preguntas de entrevista [1]
Pregunta 1
implementing K-means clustering from scratch and handling empty cluster edge cases
three rounds, each has coding + ml basic + resume related questions
understand all the details in resume is important, since might go very deep down the project you have worked on