Linkedin | Software Engineer, Machine Learning Interview Experience
Summary
I went through a 5-round interview process for a Software Engineer role at LinkedIn. The experience included a mix of coding challenges, system design, and AI fundamentals. I received three offers in the end, including one from LinkedIn.
Full Experience
As someone with 3+ years of experience as a DS/MLE in a product-based company and a B.Tech from IIT (non-circuital), I went through a multi-round interview process for a Software Engineer role at LinkedIn. The process started with a screening round that included a DSA question, a probability question, ML fundamentals, and questions on Transformers. The full loop consisted of five rounds: an AI coding round with two problems, a DSA round with two problems, an AI fundamentals/data mining round, an AI system design round, and a hiring manager round with behavioral questions.
Interview Questions (6)
Transformation and Sorting Problem
The first problem required explaining and solving a mathematical transformation followed by sorting. The second problem was about rejection sampling, similar to the rand7() → rand10() problem.
Two-Pointer Implementation Problem
This problem involved using a two-pointer technique to solve a specific array manipulation task.
Rejection Sampling Problem
The problem required generating a random number between 1-10 using a function that generates numbers between 1-7, similar to the rand7() → rand10() problem.
Partition Array into K Subsets with Equal Sum
The problem required partitioning an array into K subsets with equal sum, which is a classic backtracking problem.
Cold-Start Recommender System
The system design problem focused on creating a cold-start recommender system, discussing data preparation, feature engineering, handling sparse/implicit data, model training, and candidate generation.
Recommendation and Ranking System
This system design problem involved designing a recommendation and ranking system, covering aspects like data pipelines, feature engineering, and model deployment.
Preparation Tips
The candidate prepared by focusing on DSA, ML fundamentals, and system design concepts. They practiced coding problems, reviewed transformer models and ML techniques, and worked on system design projects to prepare for the interview rounds.