George Isichos

Fast, private inference on Apple Silicon. Behavioral limits of LLMs in sensitive contexts. Local agentic systems, so capable AI doesn't require the cloud.