The PDF Question and Answer System Using LLM and FAISS is an AI-powered document intelligence platform that transforms static PDF documents into interactive knowledge sources.
Users can upload PDFs, ask natural language questions, and receive context-aware answers generated through the combination of Semantic Search, Vector Embeddings, and Google Gemini LLM.

πΉ Upload and analyze PDF documents
πΉ Semantic search using FAISS
πΉ Context-aware answers using Gemini AI
πΉ HuggingFace embeddings for document understanding
πΉ Multi-turn conversational memory
πΉ Real-time chat interface with Streamlit
πΉ Export conversation history as PDF
πΉ Fast vector similarity search
πΉ Research-paper and technical-document support
| Category | Technology |
|---|---|
| π Language | Python |
| π¨ Frontend | Streamlit |
| π€ LLM | Google Gemini |
| π§ Embeddings | HuggingFace MiniLM |
| π Vector Search | FAISS |
| π PDF Processing | PyPDF2 |
| π€ Export | FPDF |
| π AI Domain | Generative AI, NLP, Semantic Search |
WhatsApp.Video.2026-06-10.at.13.04.09.1.1.mp4
- Mudabbir Naragaddi
- Esha Yalagi
- Omkar Yametkar
- Dhulesh Shivakale
- Prof. Savita Bagewadi
2025 IEEE 4th International Conference for Advancement in Technology (ICONAT 2025)
10.1109/ICONAT66879.2025.11362602
-
π IEEE Xplore https://ieeexplore.ieee.org/document/11362602
-
π Semantic Scholar https://www.semanticscholar.org/paper/PDF-Question-and-Answer-System-Using-LLM-and-FAISS-Naragaddi-Yalagi/185e7ef26c1b99161278e774aab1b11c1a6dbeec
π Upload PDF
β
βΌ
π Extract Text
β
βΌ
βοΈ Text Chunking
β
βΌ
π§ Generate Embeddings
β
βΌ
π Store in FAISS
β
βΌ
β User Query
β
βΌ
π Retrieve Relevant Chunks
β
βΌ
π€ Gemini AI Response
β
βΌ
π€ Export Chat History
- π Academic Research
- βοΈ Legal Document Analysis
- π’ Enterprise Knowledge Management
- π Technical Documentation
- π Educational Assistance
- π¬ Research Paper Exploration
B.E. Computer Science (Artificial Intelligence) KLE Technological University, Belagavi