3 The project portfolio and the take-home exercise

 

This chapter covers

  • How and when to invest time in your project portfolio
  • Solving take-home exercises in the most effective way
  • When and how to skip project portfolios and take-home exercises

The best way to predict the future performance of job candidates is to observe their past performance. The second-best way is to simulate a work situation and see how the candidates perform. Because the first option is not feasible (unless you are applying for a job in the same company you've worked at for a while), most of the selection processes you will encounter will resort to the second approach.

There are two ways to simulate an actual work situation. The company leads one: the take-home exercise. The candidate leads the other: the project portfolio.

A data science project portfolio is a collection of your best work as a data scientist. It includes code, visualizations, and your own commentary. For example, it could consist of problems you solved in your bootcamp, projects you assigned yourself, or actual work applications (as long as your former employers don't mind you sharing these).

In a take-home exercise, the hiring committee gives you a dataset, a problem to solve, and sometimes a series of questions. They usually ask you to solve the problem and prepare a presentation for the interview panel that discusses your approach and solution.

Let's look at both in detail, focusing on how to use them to your advantage.

3.1 The data science project portfolio

3.1.1 Selecting a topic, dataset, and skills to showcase

3.1.2 Areas a project should cover

3.1.3 Clear readme file

3.1.4 Folder structure and file naming

3.1.5 Clarity: Naming, comments and mark up, hyperlinking

3.1.6 Advanced features

3.2 The take-home exercise

3.2.1 Clarify the question to exhaustion

3.2.2 Quick EDA to identify issues to raise

3.2.3 Prioritizing work toward a solution

3.2.4 Presentation

3.2.5 Breaking the script

3.3 Using AI tools to make you more efficient

3.3.1 Baseline project research

3.3.2 Stack overflow substitute

3.3.3 Summarize for readme.md

3.3.4 Variable naming

3.3.5 Quality Control

3.3.6 AI solving attempt

3.3.7 Exercise summary

3.3.8 Presentation structure

3.3.9 Presentation slides

3.4 Summary