Installation Guide

Prerequisites

Before you touch the code, make sure you've got these installed. Nothing fancy here.

• Python 3.9 or higher — 3.10 or 3.11 is what we'd recommend. Grab it from python.org if you don't have it.
• pip — comes bundled with Python, so you're probably fine already.
• Git, optional, only if you want to clone instead of downloading a ZIP.

Step-by-Step Setup

Step 1 — Get the project onto your machine

Clone it with git clone <your-repo-url> and then cd "ai job salary". Or, if Git isn't your thing, just download the ZIP and extract it, then open a terminal inside that folder.

Step 2 — Create a virtual environment

Run python3 -m venv venv. Then activate it — the command's different depending on your OS:

source venv/bin/activate          # macOS/Linux
venv\Scripts\activate             # Windows CMD
venv\Scripts\Activate.ps1         # Windows PowerShell

Step 3 — Install the dependencies

Run pip install -r requirements.txt. This pulls in Flask, pandas, NumPy, scikit-learn, XGBoost, and joblib — give it a minute, XGBoost can take a bit to install.

Step 4 — Train the model

This is the step people skip and then wonder why predictions don't work. Run jupyter notebook ai_salary_training.ipynb and execute all the cells. When it's done, check that these three files exist:

models/salary_predictor_pipeline.joblib
models/model_metadata.json
data/ai_jobs_salaries_clean.csv

The app will actually still start without them — you'll just see a "model not found" message on /predict and a fallback notice on /visualize instead of real data.

Step 5 — Run the app

Run python3 app.py. Open http://127.0.0.1:5000 in your browser and you should see the home page with live stats and the ticker running.

Common Issues

If the notebook throws an error about a missing package, it's almost always because you ran it outside the virtual environment — make sure venv is activated in the same terminal you launch Jupyter from.

If /predict keeps showing "model not found" even after training, double-check the three files landed in the exact folders listed above — a misplaced joblib file is the usual culprit.

How to Verify It's Working

Go to /predict, fill in the eight fields, and submit. You should get a predicted salary number along with two charts underneath it — a comparison bar and a trend line. If those render, you're good.