Customer Support AI Classifier

A small machine learning style project for classifying customer support messages into categories.

This project uses a simple Naive Bayes classifier written in Python. I kept the model implementation readable because the goal is to show the fundamentals: loading data, training, predicting, and evaluating.

Categories

• billing
• technical
• account
• delivery

Tech Stack

• Python
• CSV data
• Naive Bayes classification
• basic evaluation metrics

Run

python main.py

The script trains on the sample dataset, evaluates on a small test split, and prints predictions for a few example messages.

What I Practiced

• text classification basics
• probability-based classification
• train/test split
• simple accuracy evaluation
• writing ML logic without hiding everything inside a library