explore by scenario

The first human-powered vetting for real-world Data Science πŸ‘¨β€πŸ’»

Why automated vetting of Data Scientists doesn't work

Data Science is much more than programming. It’s a spicy mixture of programming, statistical reasoning, the ability to quickly learn a business domain, and good old-fashioned problem-solving. Automated code quizzes don’t cover the important parts of the job. So companies have been forced to build their own Data Science exercises and projects in-house.

Woven’s team of experienced engineers evaluates every submission. That’s why we’re the only assessment that can cover messy, end-to-end Data Science scenarios for you. Scroll below for two examples. πŸ‘‡

"As a data scientist myself, I have gone through many assessments in the past and I found yours to be particularly innovative. The focus on evaluating a candidate's potential to solve real-world problems, rather than their ability to complete mundane tasks, sets it apart."
Woven Candidate
πŸ•™ Time Limit: 35 min

Data Wrangling: Product Usage Drop-Off

For this scenario, the candidate must use:

  • Python: The number 1 programming language for Data Science work
  • Pandas: A very common library used by Data Scientists to perform various Data Science tasks in Python

Scenario problem:

Usage is dropping in your project management product, Mello. A colleague has a hunch about user location. They need your help to create repeatable analysis. Given some CSV data, write a script that generates reports based on certain criteria.

What you’ll learn about the candidate πŸ’‘

  • Can they use Python to correctly answer business questions about a data set?
  • Do they know how to manipulate data sets and handle missing values?
  • Are they proficient with grouping/filtering/querying pandas dataframes?
  • Can they work with unclearly labeled data sets?
  • How do they structure their code: extracting helper functions or putting pre-processing in a separate step? Do they improve code quality by using descriptive variable names and what/why code comments?
πŸ•™ Time Limit: 50 min

Data Analysis: Product Usage Drop-Off

For this scenario, the candidate can pick from a variety of tools to complete the analysis including:

  • Pandas, NumPy, and Matplotlib (included in Qualified)Β 
  • Microsoft Excel
  • PowerBI
  • Jupyter Notebook
  • R

Scenario problem:

Usage is dropping off in your project management product, Mello. A coworker wants to know why. Given a CSV data file with some product usage data and a request for data analysis, write an email to a coworker to describe your findings.Β 

What you’ll learn about the candidate πŸ’‘

  • Can they visualize data to generate insights?
  • Do they explore various explanations for what could be going wrong?
  • Do they make recommendations for how to detect problems going forward?
  • How do they communicate written business insights to a coworker?
  • Can they adapt their language to effectively communicate with less technical teammates about data?


🧩 Woven fits where your code quiz or take-home fits.

The typical Woven customers saves 10.5 hours of engineering time per hire AND retains 96% of their new hires.

If this doesn’t match your needs try Woven Inbound.
Hiring data scientists?

Try Woven with real candidates for free.

Sign Up