Voice Notes — AI Transcription

Desktop transcription app with AI-powered speech-to-text, hotkey recording, and real-time processing.

Role

Solo Developer

Timeline

1 Month

Stack

Tauri, React, TypeScript

Focus

Desktop Engineering

The Challenge

Building a native desktop app with web technologies that feels truly native — handling system-level audio capture, global hotkeys, and real-time streaming transcription with minimal latency.

Architectural Decisions

◆Rust Backend

Tauri + Rust for native performance, small binary size, and direct access to system audio APIs.

◆Streaming Pipeline

Real-time audio streaming to Deepgram with chunked processing for instant transcription feedback.

◆Global Hotkeys

System-wide keyboard shortcuts for start/stop recording without switching window focus.

◆Clipboard Integration

One-click copy of transcribed text directly to system clipboard for seamless workflow integration.

Impact & Results

<500ms

Transcription Latency

15MB

App Size

12+

Languages Supported