add-voice-transcription

To transcribe WhatsApp voice notes into text using OpenAI Whisper, you configure a transcription config with your API key, install the openai npm package, and integrate the logic to download, transcribe, and store the audio content as text.

Converting WhatsApp audio messages into readable text context involves downloading the voice note, sending the audio to the Whisper API for transcription, and storing the resulting text so the agent can reference it during conversations.

Yes, you need an OpenAI API key to enable voice transcription for FFT_nano. You must add your key to a .transcription.config.json file in the project root alongside settings for the whisper-1 model before starting the application.

Yes, you can use Whisper transcription in registered WhatsApp groups to let your agent respond to audio. The transcription logic applies to voice notes sent within these registered groups, converting speech to text context.

Setup steps to add voice transcription include running npm install openai, creating a .transcription.config.json file with provider, model, apiKey, and enabled fields, then building and starting FFT_nano to process voice notes in registered groups.

Your agent ignores WhatsApp voice notes in chats because voice transcription is not configured. You must install the openai package, set up the transcription config with a valid API key, and ensure the group is registered for the agent to read audio.