PDF Knowledge Base
Upload any PDF to instantly create a searchable knowledge base.
Natural Language Q&A
Ask questions and receive contextually relevant answers.
Open Source
Easy to set up for personal or professional use.
AskLegal: PDF Chatbot with FAISS & LLaMA3
AskLegal is an AI-powered chatbot that enables users to interactively ask questions about the contents of PDF documents. By leveraging advanced language models and vector-based retrieval, it provides context-aware, document-specific answers through a user-friendly web interface.
Key Technologies/Skills
- Python & Streamlit (web UI)
- FAISS (Vector Similarity Search)
- LLaMA3 via Groq API
- LangChain (Orchestration)
- Hugging Face Embeddings
- pypdf (PDF Processing)
- sentence-transformers
What I Learned
- Designed an end-to-end pipeline for document ingestion, vectorization, and conversational retrieval.
- Integrated multiple state-of-the-art AI tools (FAISS, LLaMA3, LangChain).
- Gained hands-on experience with vector databases and LLM APIs.
- Explored best practices for error handling and user feedback in AI applications.
Notable Features
Future Improvements
- Support for querying across multiple PDFs.
- Enhanced UI interactivity and customization.
- Option to switch between different LLM providers.
Project Highlights
- Real-world application of RAG techniques.
- Integration of modern AI APIs and tools.
- Focus on usability and extensibility.