An AI-powered smartphone recommendation system built using Flask, scikit-learn, and Waitress. It takes natural language queries from users like "Best camera phone under 1 lakh" and returns personalized smartphone suggestions based on their needs.
- 🔍 Understands natural language queries
- 📊 Uses machine learning (Logistic Regression + TF-IDF)
- 📱 Recommends top 5 phones based on features (camera, battery, performance, ruggedness)
- 🌐 Provides a REST API (
/predict) for frontend consumption - 🧠 Pre-trained ML model (model.pkl) with custom vectorizer (vectorizer.pkl)
- ⚙️ Production-ready using Waitress WSGI server
| Technology | Purpose |
|---|---|
| Flask | Web API framework |
| scikit-learn | Machine Learning model training |
| pandas | CSV handling and data processing |
| TfidfVectorizer | Text feature extraction from user query |
| LogisticRegression | Query classification |
| Waitress | Production-grade Python WSGI server |
| Flask-CORS | Handles CORS for frontend API calls |
| JavaScript (AJAX) | Used in frontend to communicate with API |
git clone https://github.com/yourusername/phonefinder-backend.git cd phonefinder-backend
pip install flask flask_cors waitress pandas scikit-learn joblib
python model/train_model.py
python api/app.py
open index.html with live server (Now you can chat with this model)