$ release --version 1.1.1 --date 2026-06-16
- - Speech with Me has moved to its new home at undo.tech — this update reconnects the app to the new services.
- - Faster spoken replies and reliability improvements.
SM
// PRODUCT_DOSSIER
Speech with Me
A production-grade, fully offline AI speaking coach — on-device speech recognition, a local LLM, and neural text-to-speech in one signed, cross-platform desktop app.
LATEST: v1.1.1
Speech with Me is a full-stack, privacy-first desktop application that runs an entire AI language coach on the user's own machine — with zero cloud dependency. An Electron front end drives a bundled Python/FastAPI backend that orchestrates on-device speech recognition (MLX Whisper plus live-streaming Apple Speech), neural text-to-speech (Kokoro via MLX-Audio), and a local large language model into a natural, real-time spoken conversation — complete with a tunable coach persona, CEFR-leveled practice, session memory, and multi-language support. It ships as a code-signed, notarized, cross-platform build hardened with Ed25519-signed remote update and kill-switch verdicts, anti-rollback protection, and a PKCE desktop sign-in, alongside an optional QR-paired phone companion. End to end, it shows ownership of a real product — from low-level real-time audio and on-device ML all the way through security, release engineering, and UX polish.
Highlights
- 100% on-device AI — speech-to-text, a large language model, and neural text-to-speech all run locally; conversations, audio, and transcripts never leave the machine.
- Real-time voice pipeline — streaming microphone capture with RMS-energy / voice-activity endpointing and anti-hallucination transcript handling across MLX Whisper and live Apple Speech.
- Pluggable AI engine — one coaching core drives six interchangeable backends: Embedded MLX, llama.cpp (GGUF), LM Studio, Ollama, OpenAI, and Gemini.
- Low-latency neural TTS — Kokoro via MLX-Audio and native macOS voices, with sentence-pipelined synthesis and engine warmup so the coach starts speaking sooner.
- Full-stack desktop architecture — Electron UI plus a bundled, supervised Python/FastAPI backend with local SQLite persistence, packaged as a single installable app.
- Production security model — Ed25519-signed remote update/kill-switch, an anti-rollback monotonic clock, a tamper-evident local store, DNS-rebinding & loopback hardening, and token auth.
- Cross-platform release engineering — code-signed and notarized macOS builds via electron-builder with hardened Electron fuses, plus Windows/Linux build paths.
- Phone companion — a QR-paired LAN web app (PWA over TLS) that shares the same coach, backed by server-side cross-turn conversation memory.
- Privacy-first accounts & analytics — optional Supabase OAuth through a PKCE loopback handoff, an owner analytics console, and a signed offline licensing scaffold; telemetry stays content-free and opt-in.
ElectronPython / FastAPIOn-device LLMMLXWhisper ASRNeural TTSReal-time AudioEd25519 SecurityCross-platformSQLiteSupabasePrivacy-first
DEPLOYMENT MATRIX
| Platform | Version | Action |
|---|---|---|
| macOS | 1.1.1 | Download |
TERMINAL LOGS / CHANGELOG
$ release --version 1.1.0 --date 2026-06-15
- - Sign in with Google to keep your account across updates and devices.
- - Messages from the developer, with optional star/text feedback.
- - Updated Core Operating Protocols — please review them on first launch.
- - More reliable startup and backend connection handling.
- - Privacy: only basic, non-content usage analytics are shared — your conversations and audio never leave your device.
$ release --version 1.0.6 --date 2025-06-05
- - Major release with Hybrid AI architecture.
- - Implemented safety protocols and guidelines modal.