SAP Data Scientist Interview Guide

The SAP data science interview consists of a multi-step process interview process. The first step is a technical live Skype coding screen. The data scientist interviewer will ask about your background and your past experience. Prepare a five-minute elevator pitch on past projects, your interest in data science, and why you want to work at SAP.

After reviewing the background and past projects, the interviewer will jump into a live coding exercise and problem. These data science interview questions on the coding part will likely be in Python and done over Coderpad.

Typically, interviews at SAP vary by role and team, but commonly Data Scientist interviews follow a fairly standardized process across these question topics.

Onsite Interview

The second phase of the SAP data science interview will be on-site. This will consist of one behavioral interview and three technical interviews. The technical interviews will span a process of white-boarding coding questions, SQL and data analysis and processing questions, and machine learning conceptual questions. Make sure to review general machine learning concepts and coding exercises.

An example SAP technical interview in machine learning would be going deep into two projects that an SAP data scientist is working on. The interviewer would describe two of their current projects and would brainstorm what techniques to improve model performance. Make sure to engage in a thoughtful discussion as this is how they measure how well they would work together with you.

We’ve gathered this data from parsing thousands of interview experiences sourced from members.

Let’s look at some examples of SAP data scientist interview questions.

• How would you design a recommendation system for amazon customers? Take into consideration that a single customer could use many devices to log-on to a single account?
• What is the Big O notation for dimensionality reduction using recursion vs dynamic programming?
• How do you deal with an unbalanced dataset?
• Given a regression model, what metrics would you use to evaluate the model?
• What’s the difference between a gaussian and rbf kernel?
• Given a string, return the first recurring character in it, or ‘None’ if there is no recurring character.
• Let’s say you’re given a huge 100 gb file that cannot be entirely read into memory at once. Write code in Python to count the total number of lines in the file.

See more Deloitte data scientist questions from Interview Query: