The code assessment has 4 questions; answering 2 was sufficient to proceed to the next round.

The coding questions were exactly what is mentioned on Glassdoor.

However, other technical questions were somewhat open-ended, and it seems my answers were not what they were looking for.

1st round: 4 LeetCode-style questions on CodeSignal (2 easy, 2 medium/hard).

2nd round: 4 interviews in total

a. ML Job Fit Interview: Asked about ML basics such as bias/variance tradeoff, how to deploy models, and reducing latency during inference.

b. ML Job Fit Interview: One coding question (find the number of primes until N) and some questions about ML model deployment.

c. Case Study: Simple quantitative analysis of a cost-benefit analysis regarding simple vs. rigorous testing of software release cycles.

d. Behavioral: Three questions about past experience:

• Describe a challenging situation you encountered.
• Describe a time when you had to prioritize one task over another.
• Describe when you changed the status quo.