Show HN: I made a Raspberry with Qwen my local car AI

A developer created CarWatch, a local AI system for cars running on a Raspberry Pi 5 that uses the Qwen 35B language model to provide autonomous features like trip summaries, dashcam alerts, and voice commands—all processing offline without cloud connectivity. The system integrates with messaging platforms to communicate like an agent, accesses the car's owner manual via RAG, and maintains self-awareness of its hardware status while being remotely updatable from anywhere. The entire stack runs on approximately €300 hardware with sustained temperatures around 65°C and requires no internet, subscriptions, or wake words.