Being able to quickly glean insightful information from spoken language is essential in today's digitally connected society. The goal of this proposed work is to perform voice sentiment analysis in real-time—a cutting-edge technology with a wide range of uses. In contrast to traditional sentiment analysis, which focuses mostly on written language, this proposed work focuses on classifying sentiment into three separate groups: neutral, positive, and negative. This classification empowers people and institutions to make educated choices, improve user experiences, and provide better real-time services. This approach entails creating a real-time sentiment analysis system that can quickly and accurately classify the sentiment expressed in voice input. This is achieved by coordinating natural language processing tools, sentiment analysis models, and speech recognition libraries. The main goals of this paper include developing a flexible system that can accommodate multiple languages – namely – English, Hindi and Telugu. Additionally, it seeks to offer real-time sentiment feedback, which is a useful function in several fields, such as social media monitoring, voice assistant technologies, market research, and customer service interactions. This paper is not primarily concerned with profound emotional or attitudinal analysis, but rather with sentiment categorization, specifically focusing on positive, negative, and neutral sentiments.