Faculty Recruiting Support CICS

Data Science Interview

Your Data Science Interview guide

Data science interviews tend to be less standardized than developer interviews, resulting in a wider range of experiences and the need for more preparation. In general, applicants are tested on their understanding of fundamentals in statistics, mathematics, programming, and machine learning (depending on the job description). Employers typically follow a process which may include a coding challenge, phone screen, technical call, take home project, and final round interviews.

Coding Challenge

Some employers start with a standard HackerRank coding challenge (start anytime, timed, autoscored, flexible schedule, no live interviewer). Others test data science related technical skills by giving a dataset and asking the applicant to build a model. There are many examples online to get a sense for the variety and level of difficulty.

Phone Screen

The interviewer, typically HR, will use behavioral questions to verify background, interests, and passion for the position. Come prepared with questions about the company and role. These first interactions are an opportunity for you to interview the company back and make sure it’s a good match for your interests.

Technical Call

This round could focus on software engineering coding problems, this time with an engineer participating via code sharing platforms. This step can also often assess knowledge of machine learning, statistics, or more specific skills such as SQL. The focus is usually not on coding ML algorithms, but rather assessing their strengths and weaknesses. Scenario-based questions help evaluate your approach and experience using different models and datasets. Depending on the job description, other key skills include Python, R, D3, Apache Spark, Apache Hadoop, Apache Pig, Apache Hive, MapReduce, NoSQL databases, Github, and Tableau. Research the company on websites such as Reddit and Glassdoor to tailor your prep for their interviews.

 

Make a mock interview appointment with CICS Careers

 

Take Home Project

Although time-consuming, this step helps you show your skills in action. Before committing your time and effort, consider asking the recruiter how you will be evaluated and what to expect as far as constructive feedback.

Starting with a dataset, you might be asked to find insights, build a model (e.g., a classifier) or even pitch a business proposition. Structure your approach and analysis from the start. When you get the assignment, read the prompt and make a list of questions and assumptions to share back with the hiring manager or recruiter. Document these assumptions by adding comments in your code. Make your code readable (use good coding practices like descriptive variable naming, modular code, and comments). Test extensively. Finally write a brief summary of your thought process in a readme.

Onsite Interviews

This series of interviews will evaluate your technical and cultural fit. You will be given more data science case studies, behavioral interview questions, and other technical challenges. Onsite interviews are similar to technical calls but you get an opportunity to interact with the company in-person. Approach a problem systematically, state your assumptions, ask clarifying questions, and discuss pros and cons of each algorithm/solution you propose.