Smart Study Companion (SSC) is a fully local, privacy first study assistant designed for university students who want fast, accurate answers directly from their own lecture PDFs without relying on cloud-based AI services or file upload limits.
This project was built to solve a real academic problem:
LLM platforms are powerful, but lecture notes are often too large, restricted, or privacy sensitive to upload.
SSC removes that limitation entirely.
- 🔒 100% Local & Secure : no data leaves your machine
- 📄 Built for real lecture PDFs (large, detailed, multi file)
- 🚫 No AI APIs, no subscriptions, no upload caps
- 🧠 Context aware answers from your own notes
- 🗂️ Multiple chat sessions, each with its own PDFs and history
Built by a university student, for university students.
- Upload multiple PDFs per study session
- Ask natural language questions based on lecture content
- Get concise, relevant answers with source attribution
- Separate chat sessions for different modules
- Chat history preserved per session (while server runs)
- Modern, animated UI with smooth UX
Backend
- Python
- Flask (REST API & session handling)
Document Processing
- PyMuPDF (fitz) : local PDF text extraction
Search & Ranking
- TF-IDF Vectorization (scikit learn)
- Cosine Similarity for relevance scoring
Frontend
- HTML
- CSS (custom dark UI & animations)
- Vanilla JavaScript
Security & Design
- Fully offline
- No external APIs
- No cloud services
- No data sharing
- PDFs are uploaded and processed locally
- Text is chunked and indexed using TF-IDF
- User questions are vectorized and compared using cosine similarity
- The most relevant section is returned as a concise answer
- Each chat session maintains its own PDFs and history
- Python 3.9+
- pip
pip install flask pymupdf scikit-learn• University lecture revision
• Exam preparation
• Large module handouts
• Privacy-sensitive academic notes
• Offline studying
⸻
• No internet connection required after setup
• No third party APIs
• No telemetry or tracking
• All data stays on your local machine
⸻
Mohammed Zuoriki
Cybersecurity Student | Aspiring Cloud Security Architect
LinkedIn: https://www.linkedin.com/in/mohammed-zuoriki-856133250/
⸻
Contributions, feedback, and ideas are welcome. Feel free to fork the repository or open an issue.