Boston Consulting Group Data Scientist Intern interview questions
based on 31 ratings - Updated Apr 1, 2026
Averageinterview difficulty
Mostly positiveinterview experience
How others got an interview
88%
Applied online
Applied online
6%
Other
Other
6%
Employee Referral
Employee Referral
Interview search
31 interviews
Viewing 1 - 5 of 31 Interviews
Boston Consulting Group interviews FAQs
Candidates applying for Data Scientist Intern roles take an average of 21 days to get hired, when considering 1 user submitted interviews for this role. To compare, the hiring process at Boston Consulting Group overall takes an average of 38 days.
Here are the most commonly searched roles for interview reports -
I applied through other source. The process took 3 weeks. I interviewed at Boston Consulting Group in Feb 2026
Interview
Coding assessment, involving basic pandas questions towards data cleaning, preprocessing, structuring, and also questions regardind ML.
It involves the need to understand pandas, scikit and numpy very well.
Totally automated via Code Signal.
I applied through an employee referral. I interviewed at Boston Consulting Group in Jan 2026
Interview
AI based video interview followed by coding round which was using codesignal. Then there was one coding interview with a case question with a person and then finally 3 interviews. All interviews were about 40 minutes. Questions were mostly case questions except for one round with a very senior member who asked more theoretical questions related to statistics and machine learning. I would advise to look up about the projects and BCG blogs written by your interviews.
Interview questions [1]
Question 1
Case questions from their past client projects
Theoretical questions like bias vs variance
One technical round with basic python/EDA/ML questions, followed by a case round, followed by 3 back to back cases. The cases were testing fundamentals, and how you respond under pressure, as well as your communication and applied data science skills
Interview questions [1]
Question 1
1. Predictive maintenance case - if you have more time how will you explain to a client how you'll improve a model
2. Random forest - if you have a noisy dataset (lots of features that don't matter) whats the probability of the 2 features of interest being present in any single tree that randomly selects 50 features out of 100