https://github.com/svilupp/ios-memsieve-app

MemSieve: Your AI Dictation Assistant
https://github.com/svilupp/ios-memsieve-app

dictation genai ios-app

Last synced: 7 months ago
JSON representation

MemSieve: Your AI Dictation Assistant

• Host: GitHub
• URL: https://github.com/svilupp/ios-memsieve-app
• Owner: svilupp
• License: mit
• Created: 2024-12-09T20:05:48.000Z (10 months ago)
• Default Branch: main
• Last Pushed: 2024-12-09T20:06:50.000Z (10 months ago)
• Last Synced: 2025-02-25T13:53:39.875Z (7 months ago)
• Topics: dictation, genai, ios-app
• Language: Swift
• Homepage:
• Size: 10.5 MB
• Stars: 2
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
![MemSieve](MemSieve/Assets.xcassets/AppIcon.appiconset/iTunesArtwork@2x.png)

# MemSieve: Your AI Dictation Assistant

MemSieve is a simple yet powerful iOS dictation app that helps you capture, transcribe, and format your thoughts with ease. Record anywhere, even offline - transcribe when you're ready to connect.

## Features

- **Smart Presets**: Pre-configured templates that automatically clean up and transform your transcriptions:

  - **Default**: Simple transcription with your personal context

  - **Bullet Points**: Automatically formats text into organized bullet points

  - **Thoughts**: Structures recordings into clear sections with headings

- **Offline Recording**: Record anywhere, transcribe later when online

- **Customizable Context**: Personalize transcription prompts to improve accuracy for names, terms, and phrases you use often

- **Import Support**: Seamlessly import recordings from Apple Voice Memos

- **AI-Powered Editing**: Highlight any text and use AI to edit, reformat, or refine your transcriptions

- **Flexible Export**: Copy, paste, email, or share your transcriptions however you prefer

## Screenshots

![Recording](assets/screenshot1a.png) 

![Transcription](assets/screenshot1b.png)

## Technical Details

- Built for iOS using SwiftUI

- Uses AVFoundation for audio recording and processing

- Integrates with OpenAI's API for transcription (Whisper) and text editing (GPT-4o)

- Supports high-quality audio formats (192kbps AAC)

- Implements smart chunking for handling long recordings (silence detection based on power level)

## Requirements

- iOS 15.0 or later

- OpenAI API key for transcription services and text editing

- Internet connection for transcription (can be done later on)

## Getting Started

1. Clone the repository

2. Add your OpenAI API key to `Config.swift`

3. Build and run using Xcode

4. Grant microphone permissions when prompted

## Usage

1. Select a preset (Default, Bullet Points, or Thoughts)

2. Tap the microphone button to start recording

3. Monitor your audio levels in real-time

4. Tap the stop button when finished

5. Add a label to your recording

6. Wait for transcription and auto-formatting to complete

7. View and edit your transcribed text in the History tab