LinkedIn Staff Data Scientist Interview

Michael recentely received a staff data scientist job offer from LinkedIn. Here is what the overall process looks like:

LinkedIn Staff Data Scientist Job Interview Summary

Candidate: Michael

How it gets started: applied on LinkedIn's website

Job level: Staff (Principal)

Year of Experience: 10 years experience 

Degree: Ph.D. in statistics

Offer: Yes

TC: ~500K USD

Location: Mountain View, CA

Interview process: 4 weeks

Preparation: 1 month

Has a job: yes

Decide to join: N/A

LinkedIn Staff Data Scientist Interview Round 1: HR chat

Introduction about the different data scientist tracks/concentrations at LinkedIn. Explained the role (machine learning), the hiring team, and business problems: fraud and privacy-related large-scale ML models.

LinkedIn Staff Data Scientist Interview Round 2: Technical screen 

1. Statistics, alternatives on A/B testing, e.g., if you absolutely can't do A/B testing, can you still evaluate how well the new feature works and how. 
2. Create a Machine Learning model to predict whether a LinkedIn member is looking for job opportunities, what variables and features you will use, how you evaluate the model performance, product question: how do you obtain training labels for this model.

LinkedIn Staff Data Scientist Virtual Onsite Interview:

Round 1: statistics and A/B testing

1. What is A/A testing? Why do we need to run A/A testing?
2. How do we determine how long to run A/B testing?
3. What is your choice of the value of statistical Power? How can we improve Power?

Round 2: product sense: a case study

1. How do you improve LinkedIn's advertising?
2. How do we improve the people you may know feature?

Round 3. machine learning

1. (Very basic question) is logistic regression a regression model?
2. Cross-validation, how to partition data into training and testing.
3. AUC and ROC curve, the trade-off between variance and bias

Round 4: leadership/behavioral questions 

1. Why do you want to join LinkedIn and leave your current job?
2. How do you get a new model implemented in production?
3. How do you handle difficult conversations and move things forward?
4. Have you ever fired someone?

Round 5. SQL and basic Python

1. SQL: Simple OUTER JOIN and counting. See more LinkedIn SQL interview questions.
2. Python: basic text processing, simple map-reduce function for word counting.

Offer

Michael received the job offer after two weeks, he is still negotiating with LinkedIn on this staff data scientist job offer.

Get mentored and become 100% job interview ready for your data scientist interview, learn directly from former FAANG hiring managers,  BOOK YOUR SESSION TODAY.