Assessing Data Scientists From Their Resumes
The initial screening process is a huge time saver if done correctly. Here are my assessment categories and what I'm looking for to decide who to interview.
How do you assess a Data Scientist? On LinkedIn this morning, I talked about the challenges associated with assessing Data Science candidates, especially from their resume or profile alone. I break assessment into basic categories:
Domain Knowledge. I gauge how successful the candidate will be with the use cases we need them to built solutions for. I assess this first because the rest doesn’t matter if the candidate is setup to fail in the actual project work.
Reliable Model Development. I evaluate the candidate’s knowledge of model reliability concepts and best practices. I look for the ability to support a model with experimentation rather than simple accuracy metrics. I want to see a focus on the model after deployment to validate functionality in the real world and continuously improve performance.