Build an audio-to-text knowledge base with search capabilities
Pixeltable lets you build audio search workflows in two phases:
Install Dependencies
Define Your Workflow
Create table.py
:
Use Your Workflow
Create app.py
:
Workflow handles transcription and embedding automatically:
Intelligent sentence splitting using spaCy:
Fast search using E5 embeddings:
Audio Transcription
Uses OpenAI’s Whisper for audio transcription:
Text Processing
Splits transcriptions into units:
Vector Search
Implements search using E5 embeddings:
You can create custom search functions with different parameters:
Process multiple audio files in batch:
You can use different sentence transformer models: