Zyra is a full-stack AI-powered news application that delivers personalized, context-aware news experiences using semantic search, AI summarization, and modern mobile UI.
It is built with React Native (Expo) on the frontend and FastAPI + PostgreSQL (pgvector) on the backend.

This project demonstrates production-ready architecture, async backend design, and applied AI (RAG-style retrieval)

• Cross-platform app (Android & iOS) using React Native (Expo)
• Modern UI with NativeWind (Tailwind-style styling)
• Dark / light mode (system-based)
• Tab-based navigation (Home, Topics, Saved, Profile)
• Offline-first experience with local caching
• Save / unsave articles
• Pull-to-refresh news feed

• Semantic search using vector embeddings (Sentence Transformers)
• AI-powered context-aware summarization
• RAG-style pipeline (retrieve → summarize)
• Supports natural-language queries over news content

• Automated RSS ingestion from multiple global sources
• Duplicate detection and de-duplication
• Article content extraction (HTML parsing)
• Category inference (rule-based NLP)
• Image extraction with multiple fallbacks
• Vector embedding generation at ingestion time

• JWT-based authentication
• Secure password hashing using bcrypt
• Protected routes with token validation
• CORS-enabled API for mobile clients

• React Native (Expo)
• Expo Router
• NativeWind
• AsyncStorage
• Axios

• FastAPI
• PostgreSQL
• pgvector
• SQLAlchemy (Async)
• JWT Authentication
• FastAPI Security (HTTPBearer)

• Sentence Transformers (all-MiniLM-L6-v2)
• Vector similarity search
• Extractive summarization (extensible to LLMs)

DATABASE_URL=postgresql+asyncpg://user:password@host:5432/dbname JWT_SECRET=your-secret-key OPENAI_API_KEY=optional

git clone https://github.com/mdShakil2004/Zyra-AI-powered-news-platform.git
cd zyra/server
python -m venv venv
source venv/bin/activate   # Windows: venv\Scripts\activate
pip install -r requirements.txt

## Run the server:
uvicorn main:app --reload

#$ News Ingestion
python scripts/ingest_news.py

## Frontend
cd app
npm install
npx expo start

• Async-first backend to efficiently handle high I/O workloads
• Vector search over keyword search for improved semantic relevance
• Embeddings generated at ingestion time to minimize query latency
• Modular architecture for better maintainability and scalability
• Offline-first mobile UX to support real-world usage scenarios

• LLM-based abstractive summarization
• User-specific feed ranking based on preferences
• Chat history persistence
• Background task scheduling (Celery / worker-based jobs)
• Push notifications
• Premium subscription features

This project demonstrates:

• Real-world system design
• Applied AI beyond basic tutorials
• Secure authentication flows
• Scalable backend architecture patterns
• Production-grade mobile UI
• Industry-level engineering practices used in modern SaaS and AI-driven products

Md Shakil
Software Engineer | Backend & AI Systems

📧 Email: [email protected]
🔗 LinkedIn: https://linkedin.com/in/mdshakil2004

• ATS-safe
• Recruiter-friendly
• Interview-ready
• GitHub-professional

If you want, next I can:

• Optimize this README for GitHub stars
• Add screenshots section
• Write a case-study style README
• Rewrite for FAANG-style portfolio

Just tell me 👍