Metis — Dispatches Technical dispatches by Darwin Chen on on-device LLM inference, edge AI runtimes, and low-latency systems architecture. https://darwin27264.github.io/blog/ 2026-10-29T00:00:00Z Darwin Chen https://darwin27264.github.io/ Architecting ofln: Zero-Cloud On-Device LLM Chat on Mobile https://darwin27264.github.io/blog/posts/ofln-on-device-llm-architecture.html 2026-10-01T00:00:00Z 2026-10-01T00:00:00Z How ofln packages Q4_0 GGUF models for private offline LLM chat on iOS and Android using React Native, llama.rn, and a curated phone-sized model shelf. Darwin Chen Metis (Μῆτις): Introducing the Architecture Blog & Systems Philosophy https://darwin27264.github.io/blog/posts/introducing-the-tech-blog.html 2026-10-09T00:00:00Z 2026-10-09T00:00:00Z Why this engineering archive is named Metis (Μῆτις)—pragmatic intelligence, local-first computing, and low-latency systems philosophy. Darwin Chen