Insurance Premium Category Prediction API

This project is built to understand how FastAPI can be used to serve a machine‑learning model through an API. The model is a simple RandomForestClassifier trained on a toy dataset of 100 rows, and the focus is on learning the end‑to‑end workflow rather than building a usefull project.

---

Overview

The API exposes a /predict endpoint that accepts user details such as age, weight, height, income, and some lifestyle factors. These inputs are converted into a pandas DataFrame and passed to the prideiction API which validates the date, performes feature engineering and tranformation and ultimately make a prediction about the Insurance Premium Category.

A lightweight Streamlit UI is included to collect user inputs and send them to the FastAPI backend.

---

Tech Stack

• FastAPI for building the API
• Uvicorn as the ASGI server
• pandas for data manupulation
• scikit‑learn for the RandomForest model
• Streamlit for the frontend UI

---

API Endpoint

POST /predict

Example Input:

{
  "age": 119,
  "weight": 79,
  "height": 1.72,
  "income_lpa": 70,
  "smoker": true,
  "city": "Delhi",
  "occupation": "government_job"
}

How to use:

Start API server

uvicorn main:app --reload

Launch Streamlit UI

streamlit run app.py