Peter Dawell and Nora Kane talk about Ollama, a tool for running open-source Large Language Models (LLMs) locally. It provides a user-friendly interface and APIs for various programming languages ​​to manage and interact with LLMs, including features such as structured outputs and function calls. One focus is on data protection through local processing of data. Instructions for installing and using Ollama on different operating systems (Windows, Linux, macOS) are also provided, along with examples of practical use in various contexts, such as chatbot development and task automation.