Whisper Setup_

Voice-to-text transcription. Send voice memos and let your agent turn them into notes, summaries, or actionable items. Perfect for hands-free input.

What It Does

Whisper enables:

• Send voice messages and have them automatically transcribed
• Convert voice memos into searchable text
• Dictate notes while driving, cooking, or away from keyboard
• Let your agent process voice input as naturally as text

Installation Steps

1. Get an OpenAI API Key (if needed)

Visit platform.openai.com, sign up, and generate an API key in your account settings.

2. Add Your API Key to .env

Store your key in the workspace .env:

3. Configure Whisper Integration

Your agent should already have Whisper support. Test it by sending a voice message (if your platform supports audio upload).

4. Test Transcription

Send a short voice memo. Your agent should respond with the transcribed text:

Agent responds: "Transcribed: 'Remember to review the project report'"

Best Practices

• Keep memos focused: Short, clear voice notes work better than rambling.
• Speak clearly: Whisper handles accents well, but clarity helps.
• Provide context: Say "add to todo:" or "remember that:" to guide the agent.

Verification

Successful setup means:

• Voice files are uploaded successfully
• Transcriptions are accurate and timely
• Your agent can parse and respond to transcribed content

Troubleshooting

• API key invalid: Verify your OpenAI key is active and has sufficient credits.
• Audio not uploading: Check your messaging app supports audio and file sizes are under limits (typically 25MB).
• Poor transcription quality: Speak more clearly or use shorter phrases.