Private LLMs and RAG Pipelines in Python

Learn to build a private Large Language Model (LLM) and Retrieval-Augmented Generation (RAG) system using Python in air-gapped environments through this 19-minute conference talk from PyCon Israel. Discover how to select efficient open-source models, configure local vector databases, and optimize retrieval processes when working with limited computational resources. Explore practical implementation strategies for maintaining data privacy and security while leveraging the power of LLMs and RAG architectures in isolated network environments. Gain insights from Red Hat software engineer Yaacov Zamir, who brings extensive experience from contributing to free software projects, as he demonstrates the technical considerations and best practices for deploying these systems without external internet connectivity.