Hi HN,
I built MirrorMate, a self-hosted AI smart mirror that runs entirely on your own hardware.
It connects to a local LLM (Ollama), supports speech (STT/TTS), embeddings, and simple memory via RAG.
The goal is to have a persistent, always-on personal assistant without sending data to the cloud.
It’s still experimental, but usable today on a Raspberry Pi or a small home server.
It connects to a local LLM (Ollama), supports speech (STT/TTS), embeddings, and simple memory via RAG. The goal is to have a persistent, always-on personal assistant without sending data to the cloud.
It’s still experimental, but usable today on a Raspberry Pi or a small home server.
A short demo video is available in the README.
GitHub: https://github.com/orangekame3/mirrormate
Feedback welcome.