Dispatches on on-device LLM inference, edge AI runtimes, and low-latency systems architecture. Favoring low-latency performance, local-first computation, and quiet systems craftsmanship.
Private on-device LLM chat with no account and no cloud. Schedule a Morning Brief for local summaries, or Translate from any app via the share sheet—weights and conversations stay on your phone.
Why name this collection Metis? Exploring the ancient Greek concept of cunning resourcefulness, pragmatic wisdom, and how it informs local-first computation, low-latency runtimes, and edge AI engineering.
How we built private offline mobile LLM chat for Android and iOS using React Native, llama.rn, and quantized C++ GGUF runtimes with real-time token/sec telemetry, RAM tracking, and on-device multimodal vision.
Try searching with another keyword or reset the category filter.