mayanksinha-dev/AirSense-AI ? reverse-engineered prompt

Build me a polished local web app called AirSense AI that predicts air quality from a few environmental inputs. I want a simple homepage with a nice premium feel, glassy modern visuals, smooth animations, and a mobile friendly layout. The main flow should let someone enter temperature, humidity, PM2.5, PM10, NO₂, SO₂, CO, proximity to industrial areas, and population density, then get a prediction for the air quality level as Good, Moderate, Poor, or Hazardous.

Please use the trained model and label encoder already in the repo, wire everything up in Flask, and make the experience feel complete. The result page should clearly show the prediction, a short explanation of what the values suggest, and some basic visual insights like charts or status cards for the pollution parameters. If anything is missing, make reasonable choices and keep it clean and easy to use. It should run locally with app.py and feel like a finished demo, not just a form. Look up current docs online if you need to.