GaganAIML/InsightArchive
0
๐ค InsightArchive: Intelligent Document Explorer
InsightArchive is a high-performance RAG (Retrieval-Augmented Generation) application designed to turn static documents into interactive knowledge bases. Built with LangChain, Gemini 1.5 Flash, and ChromaDB, it allows users to chat with their data with zero hallucinations.
๐ Key Features
- Multi-Format Support: Ingests and processes both
.pdfand.docxfiles seamlessly. - Fact-Grounded Q&A: Uses semantic search to ensure answers are pulled directly from the source text.
- Source Attribution: Every answer includes a "View Sources" dropdown to show exactly which page/chunk the AI read.
- Task Automation: Optimized for academic use cases like summarization, key-term extraction, and study note generation.
๐ ๏ธ Tech Stack
- LLM: Google Gemini 1.5 Flash
- Orchestration: LangChain 0.3.0
- Vector Store: ChromaDB
- Frontend: Streamlit
- Embeddings: Google Generative AI Embeddings (text-embedding-004)
Edit /src/streamlit_app.py to customize this app to your heart's desire. :heart:
If you have any questions, checkout our documentation and community forums.
