Intelli-Docs is an AI-powered solution designed to simplify personal document management by combining Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), and image retrieval. Built as part of my Bachelor of Engineering in Computer Engineering at Tribhuvan University, Purwanchal Campus, this project tackles the limitations of traditional document handling with features like OCR-based text extraction, intelligent search, and secure cloud storage.
- OCR Text Extraction: Extracts text from scanned images (IDs, certificates) using Pytesseract.
- Image Retrieval: Uses OpenAI CLIP for visual document search.
- Intelligent Search: Supports natural language queries via RAG + LLMs.
- Secure Storage: Cloudinary-backed document storage.
- Mobile Interface: Flutter-based mobile app for accessibility.
- Python 3.8+
- TensorFlow 2.10
- Flask & FastAPI
- Pytesseract
- Cloudinary SDK
- Flutter 3.x
- (Optional) CUDA-enabled GPU
git clone https://github.com/your-username/intelli-docs.git
cd intelli-docs
pip install -r requirements.txt- Set up Flutter (see Flutter docs).
- Add Cloudinary credentials to
config.yaml.
uvicorn app.main:app --reload # Start backend
flutter run # Launch Flutter appUpload documents (PDFs, images) and ask natural language questions like:
“Who is our deputy head of department?”
➜ “Asst. Prof. Pukar Karki”
- ✅ Team Celebration
- 🆔 ID Card Retrieval
- 📲 App Interface
The complete project report (submitted March 2025) includes implementation, results, and references.
# Fork and clone
git checkout -b feature/your-feature
git commit -m "Add your feature"
git push origin feature/your-featureOpen a Pull Request to contribute!
Licensed under the MIT License.
- Kritika Thapa
- Prashant Bhattarai
- Roshan Chaudhary
- Saurab Baral
For queries, use the GitHub Issues page.