I want to transform my Hindi news portal into a largely self-running operation that gathers breaking stories, drafts publish-ready Hindi articles, finds relevant photographs, and decides where each piece appears on the page—all before I step in for a quick editorial pass. Here is what I expect the system to handle: • Automatic news generation in fluent, reader-friendly Hindi, triggered in real time by fresh items detected online. • Smart sourcing that pulls facts from official news websites, social-media updates, and curated RSS feeds, then cross-checks them against multiple independent references so that unverified or fake reports never make it into the draft queue. • Image retrieval that pairs every story with a correctly licensed photograph illustrating the actual event if available for free. • Placement logic that weighs freshness, relevance, and category to decide which article shows first, middle, or last on my homepage and section pages. • A simple back-office screen where I can see each AI-drafted article, tweak wording or the headline when necessary, and hit Approve to publish. You may use whichever combination of NLP models, fact-checking APIs, Python, Node, scraping libraries, or workflow tools you prefer, provided the end result is dependable and easy for me to maintain without deep technical knowledge. Acceptance criteria: – At least 90 % of generated articles pass manual fact-checking after the cross-reference stage. – Images load with proper attribution and match the text contextually. – Headlines and story order update automatically once I press Approve. – The full pipeline works on my existing hosting stack and can be paused or adjusted from a single admin panel. If you have built similar editorial or news-gathering automations before, tell me which models or frameworks you used and how you handled multilingual output. I am ready to start as soon as I see a credible plan and a realistic timeline.