A web application for recording and transcribing audio notes with activity tracking.
- Node.js (v18 or higher)
- MongoDB
- npm or yarn
- Clone the repository
- Install dependencies for both frontend and backend:
cd frontend && npm install
cd ../backend && npm install-
Configure environment variables:
- Create a `.env` file
- Add your LLM API key
- Configure other necessary environment variables
-
Start the development server: ```bash npm start ```
-
Configure Activity Categories:
- Access the configuration panel
- Create custom categories for your activities
- Add relevant keywords for each category
- Set category colors for reports
-
Record Activities:
- Click the "Start Recording" button
- Speak into your microphone to log your activities
- Mention times and durations for better tracking
- View real-time transcription as you speak
- Click "Stop Recording" to finish
-
Activity Processing:
- Audio and transcript are saved
- LLM analyzes the transcript for activities
- Activities are automatically categorized
- Times and durations are extracted or defaulted
- Activities are stored for reporting
- Each recording is analyzed to identify specific activities
- Activities are automatically categorized based on content
- Time tracking with smart defaults:
- Explicit times are extracted when mentioned
- Default 15-minute duration when unspecified
- Recording timestamp used for undefined start times
- End-of-day consolidation of all recorded activities
- Activities grouped by category
- Time-based analysis of daily work
- Intelligent summarization of accomplishments
- Category-based time tracking
- Monthly activity patterns and trends
- Category-based aggregation
- Time allocation insights
- Productivity analysis by category
- Visual representations of time distribution
This application primarily supports Chrome and Chrome-based browsers due to the Web Speech API implementation.
MIT