avinashchetry-cloud/Candidate-Data-Processing-Pipeline ? reverse-engineered prompt

Build me a clean notebook based data processing pipeline for a candidate success prediction project.

I already have the raw CSV files in a data folder, and I want the notebook to load them, clean the data, handle missing values, convert any messy JSON style fields into usable columns, and do some basic exploratory analysis so I can understand the dataset better. Please also add feature engineering and data encoding so the final output is ready for machine learning later.

Keep it practical and easy to run in order, with clear sections for loading, cleaning, analysis, transformations, and saving the prepared datasets. If anything in the data needs special handling, make the notebook robust and explain the steps in simple comments. If you need to check current best practices or documentation, feel free to look them up online.

The goal is to end up with a polished preprocessing notebook that turns the raw hackathon data into model ready tables.