The interview process was fast and crisp. The HR was always in contact and responded to mails promptly.

Had an introduction call with HR and was handed a take-home assignment on testing a dataset with a given model. The assignment was fairly simple and requires basic knowledge of Docker systems.

Then had a call with the manager of the team recruiting, talking about exceptions and my past experiences.

And lastly, had an onsite 3-hour (3 rounds) interview:

1. Talk on the take-home assignment, on my coding style, and how I would have done things differently.
2. Projects, expectations, and situation-based questions like best project, how I would communicate a discrepancy, and such.
3. ML pipeline design questions.

Given an NLP model, how would you design metrics for the ML pipeline?