In this competition, you are challenged to predict the **gravity** of road accidents—classifying them as either **Lethal** or **NonLethal**. You'll work with a challenging and realistic dataset that has been intentionally altered to mimic the messy, inconsistent data that data scientists often face in the real world. Your task is to clean, merge, and engineer features from four different sources to build a powerful predictive model.

This is more than just a modelling challenge; it's a test of your full data science skill set, from data wrangling to insightful feature engineering, and finally model optimization.

As a bonus challenge we want to see your best notebooks using Etiq Copilot ([Quick-Start Guide](https://docs.etiq.ai/quick-start)); not only will it give you a headstart in the challenge by effortlessly understanding your data and code, but we will also be awarding a secondary prize for the “best use of Etiq”. Here we want to see you make great use of the data science copilot and avoid wasting your time with data pre-processing! (Bonus points for any bugs found).