features

FluidVoice Feature Backlog

🔧 User Interface & Configuration

• Vocabulary UI Editor - Graphical interface for managing vocabulary replacement configuration

🎵 Audio Features

• Audio History Enhancement - Improved audio file management and playback features

📋 Backlog Features

• Developer Config File - JSON-based configuration with UI sync for developer workflows
• Audio Ducking During Recording - Automatically reduce background audio (Spotify, etc.) during voice recording

✅ Completed Features

• Parakeet-Only Architecture ✅ DONE - Radical simplification to Parakeet-only: fastest, most private, least complex transcription
• Parakeet Onboarding Streamlining ✅ DONE - Automatic Parakeet setup wizard for zero-config first launch experience
• Microphone Device Selection ✅ DONE - UI for selecting and switching between audio input devices
• Audio Level Metering ✅ DONE - Real-time audio waveform visualization in mini recording indicator with <5ms latency
• Uh Sound Removal ✅ DONE - Automatically remove filler sounds like 'uh', 'äh', 'um' from transcriptions
• Fast Vocabulary Correction ✅ DONE - Ultra-fast privacy-first vocabulary correction (150x faster than LLMs)
• Custom Vocabulary Feature ✅ DONE - Enhanced vocabulary customization with LLM-based semantic correction
• Custom Vocabulary GPT5 ✅ DONE - Ultra-fast offline vocabulary correction using Aho-Corasick pattern matching
• Miniwindow Recording Indicator ✅ DONE - Small floating window with waveform during recording (WhisperFlow-inspired)
• Parakeet Performance Optimizations ✅ DONE - Advanced optimizations for faster local transcription
• Unified Audio Architecture ✅ DONE - Streamlined audio processing pipeline
• Fn Key Hotkey Support ✅ DONE - Enable Function key combinations for global shortcuts
• Express Mode: Background Recording ✅ DONE - Revolutionary WhisperFlow-style background recording with hotkey start/stop
• Parakeet v3 Multilingual Upgrade ✅ DONE - 25 European languages with automatic detection and performance boost
• Performance Metrics & Language Detection ✅ DONE - Comprehensive transcription benchmarks and German language fix
• Model Preloading & Streaming UX ✅ DONE - PreloadManager system eliminates first-use delays
• WhisperKit Preload System ✅ DONE - App-idle preloading with warmup cycles
• Hybrid Build System ✅ DONE - 4.7x faster development builds

🎯 Feature Guidelines

Features in this backlog follow these principles:

• User-First: Address real user pain points identified through usage
• Technical Excellence: Maintain Swift/SwiftUI best practices
• Minimal Scope: Focus on core voice recording and transcription workflow
• Test Coverage: All features require comprehensive XCTest coverage

📁 Feature Documentation Structure

Each feature includes:

• Problem Statement - What user pain point does this solve?
• Technical Solution - Implementation approach and architecture
• Success Criteria - How do we measure completion?
• Testing Strategy - Unit tests and integration scenarios

📝 Documentation Requirements

When adding a new feature:

1. Create feature documentation in docs/features/[feature-name].md
2. Update this README - Add feature to appropriate section (High Priority, Completed, etc.)
3. Update status table - Include priority, status, and impact description
4. Link consistently - Use relative paths to feature documentation files

This ensures the feature backlog stays current and discoverable for all contributors.