I want to put our business-focused articles, landing pages, product decks, and email sequences on autopilot by combining large-language-model generation with high-quality English localization. The goal is a repeatable workflow that accepts a brief or Nepali draft, produces polished English copy that reads natively, and stores everything in our CMS with minimal human touch. Scope • Build or configure the automation pipeline (Python, JavaScript, Zapier/Make, or comparable tools) that calls OpenAI or a similar LLM, applies custom prompts, and returns publication-ready text. • Design prompt templates that preserve the original intent yet tailor style, tone, and terminology for an international business audience. • Implement guardrails for factual accuracy, compliance, and brand voice consistency. • Provide a small test suite and example inputs/outputs so I can verify quality and extend the flow later. Deliverables • Working code or no-code scenarios with clear setup instructions • Prompt library with inline documentation • Sample localized business content (at least 3 pieces) • Short usage guide covering maintenance and future language expansion Acceptance criteria Content must be fluent, culturally appropriate, and require no more than light proofreading to ship. Turnaround from input to finalized English copy should be under five minutes per item on standard hardware. If this proves effective, I may extend the same framework to creative and technical assets or add Nepali and other languages, so keep the architecture scalable.