Data Scientist

Data Scientist (Coding) qualifications to look for

Data scientists study data to help guide important decisions within a business.

Your top data scientists will gauge the complexity of a particular approach to a problem and propose alternative solutions. They take into account available time and resources and work fast on short notice to create simple but functional solutions. 

Given more time, these superstars can create elaborate designs that rival your best solutions.  

Look for candidates with strong coding and software engineering skills. They should understand the tools, processes, and exigencies for creating and maintaining software deployed to production. 

Data scientists typically hold a degree in computer science, but candidates with engineering and mathematical backgrounds can develop strengths in practical software engineering skills. 

Keep an eye out for candidates who have:

• Solid programming background in languages such as C++, Java or Scala
• Exceptional knowledge in databases
• A background with platforms that deploy machine learning solutions (Azure ML or PredictionIO, for example).
• A background with platforms like Apache Spark, Hadoop and Snowflake
  

Top tip: Hire candidates willing to grow by making sure their personal career goals align with your company's mission.

Programmer job interview questions

• What programming languages are you most comfortable working with?
• Tell me about a data science project you’ve worked on where there was a significant programming component?
• Tell me your experience with contributing to open source projects.
• Do you have any experience with online platforms for machine learning like IBM Watson Studio, Azure ML or PredictionIO?
• How would you use a logistic regression model in our company’s environment to evaluate success? Do you have experience using recommender systems for rating?
• What is hashing? Tell me when you’ve used this, or how you would use this?
• Can you tell me what dynamic programming is and how it’s used? What about recursion?

Software engineer interview questions

• What is your experience in working within a team that follows a particular agile process?
• How do you manage technical debt with data-driven models in real-world deployment? 
• Tell me how you have tested code at previous positions. Do you write tests, and if so, what kind?
• Describe how you would monitor a model over time to ensure performance doesn’t degrade?
• How would you keep a record of computations your computer model performs during production?
• How do you manage version control for top efficiency and clarity?
• Can you tell me what software patterns are, and with which patterns you are familiar? Describe when you would decide to use a specific pattern.
• Can you define PMML and what a training environment means?

Computer data interview questions 

• Have you used technology from the Hadoop stack (e.g., MapReduce, Pig, Hive, etc.)?
• Take me through how you’d take a large collection of documents and build a search engine for that collection?
• Have you worked in distributed environments, and if so, which ones?
• When using a map-reduce paradigm, what does the map function do? What does reducing do? What about the combiner and partitioner?