Siesta joins the Macs your team already owns into one private, OpenAI-compatible endpoint. Run any Ollama model, decisions, embeddings, batches and coding agents on whichever Macs nobody is using. When someone sits down, their Mac steps aside.