100% Locally Running AI-Powered Knowledge Retrieval System
VectorMind is a fully open-source, AI-powered knowledge retrieval system that leverages state-of-the-art embedding models and LLMs. It allows users to build their own vectorized knowledge bases by uploading documents or scraping websites and then querying them using advanced AI models.
✅ Open-Source AI Stack – Uses Ollama for local LLM inference, Nomic Embed Text for embeddings, and ChromaDB for vector storage.
✅ Multi-Format Support – Upload and process PDFs, DOCX, and TXT files.
✅ Web Scraping – Extracts content from websites, including Wikipedia.
✅ Efficient Vector Search – Stores and retrieves relevant knowledge efficiently.
✅ Flexible UI – Provides both Gradio and Streamlit interfaces for ease of use.
- Vector Database: ChromaDB
- Embeddings: Nomic Embed Text
- LLM Inference: LLaMa3.2 via Ollama
- Text Processing: LangChain
- Frontend: Gradio, Streamlit
- Web Scraping: BeautifulSoup
1️⃣ Upload documents or provide a website URL
2️⃣ Convert text into embeddings using Nomic Embed Text
3️⃣ Store vectorized data in ChromaDB
4️⃣ Query the knowledge base and retrieve AI-powered answers