Rupam Creator Studio for AI Assisted Content Production and Multilingual Captioning

Abstract

Short video production requires creators to move between topic research, writing, visual generation and caption editing. Rupam Creator Studio connects these activities through a web application with reusable creator context and editable intermediate outputs. The system combines public-source research, structured script generation, thumbnail and avatar service integrations, and a multilingual caption editor. Its caption pipeline separates speech recognition, acoustic alignment, text revision and rendering so that design changes do not require a new transcription request. A persistent local worker processes caption jobs, while Supabase supports authenticated asset storage and library records. This paper describes the implemented architecture, data flow and failure handling, and reports a local regression run in which 29 software checks passed. The result supports specific functional properties, including source-link validation, revision handling and job recovery; it does not establish transcription accuracy or creative effectiveness. The contribution is an integrated, human-reviewed production workflow rather than a new generative model.