Multilingual AI News Aggregator App

I’m building a real-time, AI-powered news aggregator that fetches stories from leading Indian newspapers and TV channel feeds, removes duplicates, and converts each article into crisp bullet-point summaries. The first release must speak to readers in Hindi, English and Gujarati, with the codebase structured so I can plug in more Indian languages later.

Key experience I’m looking for includes NLP summarisation (transformer models, GPT, BERT or similar), language detection, and high-quality text-to-speech. The read-aloud feature should trigger from any article screen and play flawlessly for visually-impaired users on slow connections as well.

The product has to feel native on iOS and Android and remain fully functional on the web; whether you choose React Native, Flutter or a well-organised trio of Swift, Kotlin and ReactJS is up to you as long as the UX stays consistent. Real-time push notifications, topic-based filtering, and an admin console for adding/removing sources are essential, and I’ll need clean, well-commented source code plus deployment scripts for AWS or GCP.

Deliverables
• End-to-end source code for mobile apps and web portal
• Backend services and crawler/scraper modules with unit tests
• Working TTS integration for Hindi, English, Gujarati
• Admin dashboard and database schema
• Installation and hand-over documentation

Acceptance criteria
• New articles appear in the feed within five minutes of publication
• Bullet-point summaries average 50–80 words, retain factual accuracy, and display bilingually when applicable
• TTS playback starts in under two seconds and continues offline once cached
• User can follow/unfollow interests, and feeds refresh accordingly without a full app restart

If this matches your skill set, let’s talk through your preferred tech stack and a realistic timeline for a working MVP.