A Multilingual Customer Support Assistant Using Machine Learning and Streamlit for Real-Time Global Communication

In this paper, a customer support language assistant is developed, which can be used through a web-based interface to enhance real-time multilingual communication. The tool leverages automated language detection, language translation, and reply to conversion across several scripts and languages through Machine Learning (ML) and natural language processing (NLP). This application combines a Naive Bayes classifier with Google Translate API, and a Streamlit based frontend that provides a cohesive user experience for anyone, regardless of their language proficiency. This assist offers a quick option for organizations that manage international client associations. This application also provides the transcript of .pdf file translation and conveying reply to its detected language. Within two seconds, translation and detection are completed, and up to 17 languages are recognised. Future developments, such adding voice- based inputs and broadening language support, should make this tool essential for international communication.