A powerful RAG-based internal knowledge base that allows employees to chat with company documents (PDFs) using hybrid search and Groq LLM.
- PDF Upload & Processing with automatic chunking
- Hybrid Search (Vector + BM25)
- AI Chat powered by Groq (Llama-3.3-70B) with source citations
- JWT Authentication with role-based access
- Uploaded Documents Management (View + Delete)
- Multiple Chat History
- Smart Query Suggestions
- Analytics Dashboard
Clean, modern dark-themed interface with sidebar navigation, document management, and interactive chat.
- Backend: FastAPI, ChromaDB, Rank-BM25, Groq
- Frontend: HTML + Tailwind CSS + Vanilla JS
- Embeddings:
all-MiniLM-L6-v2
git clone <your-repo-url>
cd Enterprise-Knowledge_Assistant
python -m venv venv
source venv/bin/activate # Windows: venv\Scripts\activate
pip install -r requirements.txt
2. Add Groq API Key
Create .env in root:
envGROQ_API_KEY=your_groq_api_key_here
3. Run
Backend: cd src && python main.py
Frontend: Open frontend/index.html with Live Server
🔑 Default Login
hr@company.com / hr123
finance@company.com / finance123
admin@company.com / admin123
📁 Project Structure
text├── src/ # FastAPI backend
├── frontend/ # HTML + JS UI
├── uploaded_docs/ # Uploaded PDFs
├── chroma_db/ # Vector database
├── screenshot.png # Project screenshot
└── README.md