Built a production-oriented multilingual speech translation pipeline for classroom lectures and technical presentations. The system processes English speech, stabilizes streaming transcripts, preserves STEM terminology, translates content into five Indic languages, and generates synthesized speech output.