Parley turns your Apple Silicon Macs, NVIDIA workstations, and spare hardware into a shared on-premise LLM cluster. Local AI inference with zero config, zero cloud costs, and zero data leaks.