Akaike AI: News Summarization & Hindi TTS App

🚀 Features

• ✅ Extracts top news articles using Serper API
• ✅ Scrapes content using BeautifulSoup
• ✅ Summarizes & analyzes sentiment using OpenAI
• ✅ Performs comparative sentiment analysis
• ✅ Converts the final summary to Hindi speech using Coqui TTS
• ✅ Streamlit UI + FastAPI backend

🛠️ Setup Instructions

1. Clone the repository:

2. Create virtual environment:

2.1 If using virtualenv:

2.2 If using conda:

3. Install dependencies:

4. Set environment variables:

OPENAI_API_KEY=your_openai_key
SERPER_API_KEY=your_serper_key

5. Run API server:

6. Run Streamlit app:

🌐 Hugging Face Spaces Deployment

• Ensure requirements.txt and all files are committed
• Push to a public GitHub repo
• Create a new Space on Hugging Face with type: Gradio/Streamlit
• Connect your repo and deploy!

📦 Project Structure

├── app.py          # Streamlit UI
├── api.py          # FastAPI backend
├── utils.py        # Utility functions
├── requirements.txt
├── .env            # API keys (not pushed)
└── README.md       # You're here

🧠 Models & Libraries

• OpenAI GPT: Summarization, sentiment, and comparative analysis
• Coqui TTS: Offline, open-source Hindi TTS
• BeautifulSoup: Web scraping
• Serper.dev: Google-like search API for news discovery

📌 Assumptions & Limitations

• News content scraping may fail for JS-heavy sites
• Hindi TTS works best with simple sentence structures
• Only 10 articles are selected for analysis

📫 Contact

✅ Submission Checklist

• app.py, api.py, utils.py, requirements.txt
• Hugging Face Space deployed ✅
• README with instructions ✅
• Proper modularity and code comments ✅