Tom Fenton moves from local AI concepts to hands-on tools for matching LLMs to hardware, running local chatbots with Ollama and benchmarking AI performance.