Background:

• 3+ YOE as DS/MLE in a Product Based Company
• B.Tech from IIT (non-circuital)

Screening Round (Eliminative)

• 1 DSA question
• 1 probability question
• ML fundamentals
• Questions on Transformers

Full Loop (5 Rounds)

1️. AI Coding Round

• Expectation: explain and solve mathematically → code it.
• First Problem on transformation + sorting
• Second problem on Rejection sampling problem (very similar to rand7() → rand10())

2️. DSA Round

• One Two-pointer implementation problem
• Partition array into K subsets with equal sum

3️. AI Fundamentals / Data Mining Round

• Deep dive into project
• Questions on Embeddings, tokenization, attention, BERT, Regularization (dropout, weight decay), Topic modeling, Clustering + ML fundamentals, etc

4️. AI System Design Round

• Problem: Cold-start Recommender System
• Discussed heavily:Data preparation & pipelines, Feature engineering strategy, Handling sparse/implicit data, Model training & candidate generation, etc

5️. Hiring Manager Round

• Introduction + behavioral questions
• One project deep dive
• A system design problem - recommendation and ranking system

Compensation details here - https://leetcode.com/discuss/post/7348795/linkedin-software-engineer-machine-learn-a5zh/

P.S. (Motivation for anyone reading this)

Don’t get disheartened after 2–3 rejections.

I’ve interviewed at Amazon, Meta, Google, CRED, Cohesity, GoDaddy, and many more. I reached the final round in most and got rejected for different reasons each time.

After 5–6 months of consistent grind, I finally ended up with 3 offers — one of them being LinkedIn.

So keep going -
As they say in Three Feet from Gold, you might be just one strike away from the breakthrough.