🎙️ Smart Audio Monitor

Smart Audio Monitor is an intelligent audio recording app designed to make capturing conversations, meetings, and key audio moments seamless and efficient. Instead of recording everything continuously, this app only records audio when it matters — when sound is detected above a set decibel threshold — and automatically stops during silence.

🚀 Features

• Smart Monitoring: Automatically starts recording when audio exceeds a certain decibel level.
• Automatic Pause & Resume: Pauses during silence and resumes when sound is detected again.
• Multiple Audio Clips: Generates separate audio files for each segment of detected sound.
• Can be used for: AI-Powered Transcription: Instantly send recordings to an AI assistant to generate:
  • Accurate transcriptions
  • Concise meeting notes
• User-Friendly Interface: Simple, clean, and focused on productivity.

🛠️ How It Works

• Start Monitoring: "Start Monitoring" is where it all starts. If you are using our index.html, just tap start monitoring button. If you want to use the npm package, create an object for SmartVoiceRecorder and call "startMonitoring" function to let the magic start.
• Smart Capture: The app listens passively and only records audio when it detects sound above the configured decibel level.
• Segmented Recordings: Each audio segment is saved individually, allowing for more organized and manageable files.
• Review Outputs: View or download transcriptions, meeting summaries, or action items derived from the audio.

📲 Getting Started

• Clone this repo or download the app package.
• Test out the features by opening index.html which has cool visualizations for the audio monitors and gives you the recordings without silences.
• Run the app on your device. Click "Start Monitoring" and let the app handle the rest!

NPM Package

• To use it in your own app, you can download the npm package and create a SmartVoiceRecorder object.
• Creating the object will automatically start an indexDB on your browser.
• You have access to methods like startMonitoring, storeChunksInDB, getRecordings and deleteRecording by id.
• You also have access to two variables rawLevel and speechLevel using which you can visualize the audio. You can receive these variables on event "audioProcess".
• By default the app stores chunks in db every 10 seconds.
• You can pass in optional parameters voiceThreshold, silenceDuration, minRecordingDuration, mimeType for customizations.
• For examples checkout our index.html.

🔧 Configuration Options

• Decibel Threshold: Adjustable to fine-tune what constitutes "sound worth recording."
• Max Silence Duration: Set how long the app waits before concluding a segment.
• Min Recording Duration: Set how short your recording could be.