LIVE DEMO: https://quickeys-tutor-e7560b.netlify.app
Quickeys is an intelligent typing assistant that uses a Hybrid AI Architecture to provide real-time coaching. Unlike standard typing tests, Quickeys analyzes why you made a mistake—whether it's speed, accuracy, or specific key patterns—and provides instant, personalized feedback using Google's Gemini 1.5 Flash LLM.
This project uses a unique two-stage AI system:
- Random Forest Classifier (Local ML): Instantly classifies user skill level (Beginner/Int/Adv) based on keystroke dynamics (WPM, Error Rate, Backspace usage) using a pre-trained
scikit-learnmodel. - Gemini 1.5 Flash (Cloud LLM): A semantic analysis engine that reads the user's typed text to generate human-like, constructive feedback (e.g., "You consistently miss capitalization on the left hand").
- Real-time Analytics: Tracks WPM, Accuracy, and Raw Mistakes instantly.
- AI Feedback Engine: Generates specific, actionable advice after every session.
- Smart Progress Tracking: Visualizes improvement over time with interactive charts.
- Secure Authentication: JWT-based login system with Bcrypt password hashing.
- Persistent History: SQLite database stores all past sessions.
- Glassmorphism UI: Modern, responsive interface built with vanilla CSS variables.
- Frontend: HTML5, CSS3, Vanilla JavaScript, Chart.js (Hosted on Netlify)(https://quickeys-tutor-e7560b.netlify.app)
- Backend: Python, Flask, Gunicorn (Hosted on Render) (https://smart-typing-tutor.onrender.com)
- AI/ML: Google Gemini API, Scikit-Learn, NumPy, Joblib
- Database: SQLite
- Security: JWT (JSON Web Tokens), Bcrypt, Dotenv
- Aditi Mehta – FullStack Development and AI/ML Implementation
- Nandana v. – UI/UX Design and Frontend Development
git clone [https://github.com/Dynamic-ctrl/Smart-Typing-Tutor.git](https://github.com/Dynamic-ctrl/Smart-Typing-Tutor.git)
cd Smart-Typing-Tutor