Scale AI Engineer in New York: seven rounds, a deck of cards LLD, a RAG pipeline OOP round and two ML rounds that ended it

I applied for a role as AI engineer based in NY in the summer. The process overall went well but definitely a tough interview process. I think it was 7 rounds total:

1. Initial recruiter call: this was the same recruiter who reached out to me on LinkedIn, asking about my past experiences, telling me about the company and role.

1. Technical call, LLD round: This question was to implement methods based on a deck of cards, shuffling, dealing, sorting and viewing the cards. Speed was definitely important here as it was time constrained.

1. Manager Call: Going through past experiences again and talking through projects I'd worked on that were on my resume.

1. Last round was 4 in one, a virtual onsite/superday with 4 interviews back-to-back:

First was technical (OOP/LLD again), based on a simple RAG pipeline.

Next was typical behavioral, an example of how you dealt with disagreement. All STAR method type questions.

Third and fourth were both ML rounds, some theory questions on ML fundamentals and then reading a research paper and answering questions based on it. I tripped up in the ML Fundamentals interview and was obviously super disappointed after dedicating a lot of time. You definitely want to know your ML theory for this.