Jae Wilson @DataCrew

// projects / mermaid

Voice → Diagram

voice input · real-time Mermaid generation · local LLM

Speak or type a process description and watch a Mermaid flowchart, sequence diagram, or Marp slide deck generate in real time. Powered by local Whisper + Ollama.

250ms

chunk interval

Whisper processes audio in real time

local

LLM backend

Ollama or vLLM — no cloud required