The interview process involved five rounds, along with frequent sync-ups with the recruiter.

Round 1: Phone Screen This involved a HackerRank coding question similar to weighted random sampling. I was required to create a test that proved the solution worked. There was a significant discussion around statistics, distribution testing, the precision of the solution, and its potential failure points. A working solution with rigorous test cases was expected.

Virtual On-site Round 1: Behavioral (Teamwork)

• Discuss a project worked on within a team, including challenges, your contribution, disagreements, and what could have been done differently.
• Explain why you are interested in Yelp and why you are leaving your present company.
• State your aspirations from Yelp.
• Describe a decision you made and how it impacted the team.

Round 2: Behavioral (Individual)

• Discuss an individual project and its impact.
• Provide details about stakeholders and the organizational structure at your present company.
• Explain a new technology you worked with, why you chose to learn it, and how it benefited you.

Round 3: ML Design This round focused on a recommendation-based problem, emphasizing problem formulation, data exploration, and machine learning modeling. The approach was to start incrementally from the simplest rule-based model and then progress to more complex deep learning-based solutions.

Round 4: Problem Solving This involved a coding question similar to n-gram generation but with customized rules and precedence. A complete working solution along with test cases was expected.

Questions on Logistic Regression and regularization.

Statistical significance and chi-square test.

Recommendation system and the cold-start problem.

Why Yelp?

What is the most important thing at the workplace?