### Interview Rounds Overview - Round 1: Phone Screen - Round 2: Hiring Manager (HM) & Behavioral (BQ) - Round 3: Coding - Round 4: ML System Infrastructure - Round 5: ML System Design

The phone screen consisted of one coding question.

### Round 1: ML System Design **Question:** Design a recommender system: Where to save data? How to serve models? What is a reasonable number of candidates to pass into the ranking model, and how is it calculated? Given a 60GB memory GPU, how to calculate the...

I had a technical phone interview. It included four multiple-choice questions related to machine learning and two coding questions that focused on NumPy.

I had a technical phone screen where I was expected to code efficiently. I failed because I didn't finish the questions in time. **Question 1:** LeetCode-style coding question - LeetCode 901. **Question 2:** Easy SQL question. **Question 3:** Medium SQL questi...

### Round 1: Coding **Question:** K Closest Points - LeetCode problem. --- ### Round 2: System Design **Question:** Design an ML job scheduler. I lacked experience in this area, and the interviewer became increasingly unimpressed. --- ### Round 3: System Desig...

### Round 1: Coding **Question:** A variation of LeetCode 332 with a follow-up to print all paths. --- ### Round 2: Problem Solving **Question:** Open-ended question: Design a movie recommendation system.

I had a technical phone screen where I was asked to implement a function that prints events, but only once within a 1-minute window. The input events have a name and a timestamp.

I implemented a weighted cache with `get` and `put` operations. The cache has a total weight limit. When the limit is reached, the key-value pair with the largest weight is removed.