AudioScribe

Audio Content Transcription Platform

A streamlined solution designed to transform podcasts, YouTube videos, and audio content into accurate, searchable text. AudioScribe makes audio content accessible, searchable, and more valuable through automated transcription.

Demo

Try it live at: AudioScribe Web App

Overview

Description: Educational project focused on building a platform that converts various audio sources into text transcriptions, leveraging multiple content provider APIs and AI-powered transcription services.

Challenge: Learning to integrate and orchestrate multiple third-party APIs (ListenNotes, Spotify, YouTube, AssemblyAI) into a cohesive application while managing different data formats, authentication methods, and providing a seamless user experience.

Tech Stack

• 
• Various Content APIs (ListenNotes, Spotify, YouTube)

Key Features

Multi-Source Support:

• Podcasts from ListenNotes and Spotify
• Videos from YouTube
• Direct audio file uploads

Advanced Processing:

• Speaker identification
• Automatic language detection
• Content summarization

Installation

1. Clone the repository:

git clone https://github.com/jbo-tech/audioscribe.git
cd audioscribe

2. Install dependencies:

pip install -r requirements.txt

3. Configure API keys: Create a /.streamlit/secrets.toml file with:

listennotes = "your_listennotes_api_key"
assemblyai = "your_assemblyai_api_key"
[spotify]
id = "your_spotify_client_id"
secret = "your_spotify_client_secret"

4. Launch the application:

streamlit run transcrypt-app.py

API Integration

The project integrates with several third-party services:

• AssemblyAI for audio transcription
• ListenNotes for podcast content
• Spotify API for podcast access
• YouTube for video content

Acknowledgments

• Thanks to ListenNotes, Spotify, and YouTube for their API access
• Thanks to AssemblyAI for providing transcription services

---

Last Updated: 26/10/2024