Select Your Challenge

This is an opportunity to show your skills! These problems are not meant to be overly challenging and participants don't need previous machine learning or data science experience to complete them successfully. Your approach to these challenges helps us understand your capabilities in problem solving and rapidly acquiring and demonstrating new skills.

Pick one of the following challenges. Perform your analysis in a Google Colaboratory Notebook and share your results with us! 

Some hints for hacking our challenge:

  • Ask yourself, why would they have selected this problem for the challenge? What are some gotchas in this domain I should know about?

  • What types of visualisations will help me grasp the nature of the problem / data?

  • What are some of the weaknesses of the model and and how can it be improved with additional work?

  • Your challenge will be an important part of your interview; be prepared to discuss your work with us!

You can learn more about Google Collaboratory here.


Linear Model

Fit a regression model (with an out-of-sample R^2 of >0.01) to a financial time series. Financial series have a low signal-to-noise ratio, so even a weak correlation will require some basic feature engineering (moving average prices, volatility, etc).

Basic Computer Vision

Modify the Keras MNIST example to train a single decision tree model.  Show the training / validation accuracy of your model.

Recommender System

Build a movie recommender system using a Python library such as Surprise or FastFM.


Share Your Challenge


Once you are ready to submit your results, please follow these steps:

Step 1.  Click 'SHARE' on the right hand side of your Collaboratory notebook


Step 2. You should see the following window appear. Under the 'People' section, add and click 'Send'. That's it!

Screen Shot 2018-03-12 at 14.57.47.png

Note:  If you have any problems sharing your notebook from Colab, please download the .ipynb file (File -> Download .ipynb) and email it to