5 data science questions a hiring manager must ask
https://medium.datadriveninvestor.com/5-data-science-questions-a-hiring-manger-must-ask-49b8bb512228
Over time, I have come to rely on these five questions for assessing if a data scientist could be a good fit.
**I start with a real-world case study and ask about how the candidate will solve various problems related to that case study. **AI or Deep Learning is not the answer to every problem. In fact, most of the situations can be solved with much simpler approaches. (See this)
- **Then I pick a problem, esp. if the candidate suggests using some neural network, and ask the candidate to estimate how much data is necessary to train the model for this problem. **This lets me assess whether the candidate knows basic algorithms, and also whether they have implemented an AI model before. This discussion inevitably goes into modeling, data structure, etc. as well.
- **Then I ask if the data available is only 1/2 or 1/10th (based on situation) of what the candidate wants, what would he/she do? **This lets me assess the candidate’s grasp of basic principles, their creativity, and their understanding of the real-world situations. In the real-world, very few enterprise teams all the data they need. (See this answer).
- In most cases question 3 results in a solution which is not as accurate as expected in the case study. I ask the candidate how they further improve the accuracy of acceptable standards. It is important to understand that an AI model is not a standalone be-all-end-all. In most practical applications, AI must work with humans, heuristics, and other validation processes/ sources to give acceptable results.
- **Finally, I ask the candidate about the ROI of this process in the case study. **I don’t expect a data science candidate to know about all the non-technical aspects of a business, but it is important for them to realize that this is not an isolated research project — it must live in real-world solutions. (See here).