post-flow

YouTube Transcription and Summarization Service

This project provides a service that searches for YouTube videos based on a query, downloads the audio, transcribes the audio using a third-party API, and then summarizes the transcriptions using ChatGPT API. It's built using Go and integrates various external services for transcription and summarization.

Features

• YouTube Video Search: Search for videos on YouTube using a specific query.
• Audio Extraction: Download and extract audio from the YouTube videos.
• Transcription: Send the audio to a transcription service and receive text transcriptions.
• Summarization: Summarize the transcriptions using ChatGPT API.

Setup and Installation

1. Clone the Repository:

   git clone https://github.com/GabrielDillenburg/post-flow/tree/main
   

2. Set Environment Variables:

   • YOUTUBE_API_KEY: Your API key for the YouTube Data API.
   • TRANSCRIPTION_API_URL: The URL for the transcription service.
   • TRANSCRIPTION_API_TOKEN: The Bearer token for the transcription service.
   • CHAT_GPT_API_URL: The URL for the ChatGPT API.

3. Install Dependencies (if any):

   go get ./...
   

4. Run the Service:

   go run main.go 
   

   or

    podman-compose up
   

Usage

Send a POST request to /transcribe endpoint with a JSON body containing the search query. For example:

{
  "query": "your search term"
}

The service will search YouTube videos, download and transcribe the audio, and return a summarized response.

Project Structure

• server.go: Initializes the HTTP server and defines endpoints.
• youtube-transcriptor.go: Handles the logic for fetching YouTube videos.
• downloader.go: Manages the downloading and extraction of audio from YouTube videos.
• processor.go: Orchestrates the processing workflow, including transcription and summarization.
• summarizator.go: Handles the communication with the ChatGPT API for summarization.
• transcriptor.go: Manages the transcription requests to the third-party API.

Contributing

Contributions to the project are welcome! Please create a pull request with your proposed changes.

License

This project is licensed under the MIT License - see the LICENSE.md file for details.