An interactive AI-powered podcast co-host that can engage in natural conversations, process context from documents, and respond with voice.

• 🎙️ Voice Interaction

  • Real-time voice recording
  • Speech-to-text using OpenAI's Whisper API
  • Text-to-speech using ElevenLabs API

• 📚 Document Processing

  • PDF upload and processing
  • Vector storage using FAISS
  • Context-aware responses

• 🤖 AI Conversation

  • GPT-4 powered responses
  • Memory of conversation history
  • Natural, podcast-style interaction

• 💻 Modern Interface

  • Clean UI with Tailwind CSS
  • Real-time audio visualization
  • Easy-to-use controls

• Backend: FastAPI (Python)
• Frontend: HTML, JavaScript, Tailwind CSS
• AI/ML:
  • OpenAI GPT-4 & Whisper
  • ElevenLabs TTS
  • LangChain
  • FAISS Vector Store

1. Clone the repository:

   git clone https://github.com/yourusername/ai-podcast-cohost.git
   cd ai-podcast-cohost

2. Install dependencies:

   pip install -r requirements.txt

3. Create a .env file in the project root:

   OPENAI_API_KEY=your_openai_api_key
   ELEVENLABS_API_KEY=your_elevenlabs_api_key
   

4. Run the application:

   python app/main.py

5. Open your browser and navigate to http://127.0.0.1:53997

1. Upload Context (Optional)

   • Click "Upload PDF" to add background knowledge
   • The system will process and index the content

2. Start Conversation

   • Use the voice recorder: Click "Start Recording" to speak
   • Or type messages in the text input

3. Interact with AI

   • View your transcribed speech
   • See AI responses in text form
   • Listen to AI responses through the audio player

ai-podcast-cohost/
├── app/
│   ├── static/
│   │   ├── uploads/    # Temporary file storage
│   │   └── audio/      # Generated audio files
│   ├── templates/
│   │   └── index.html  # Main UI template
│   ├── main.py         # FastAPI application
│   └── utils.py        # Core functionality
├── requirements.txt
└── .env

Contributions are welcome! Please feel free to submit a Pull Request.

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