Jarvis: AI-Enhanced Desktop Virtual Assistant

This paper presents the development and evaluation of “Jarvis: AI-Enhanced Desktop Virtual Assistant,” a multi- functional system designed to automate daily tasks through voice commands and natural language processing (NLP). Leveraging advanced technologies such as OpenAI's GPT-4, machine learning models, Optical Character Recognition (OCR), and web automation tools, Jarvis integrates voice, text, and graphical interface interactions to deliver an intuitive user experience. With a 95% accuracy in voice command recognition and a 98% task completion rate, Jarvis efficiently handles diverse operations, including code generation, web scraping, and GUI automation. Key functionalities are supported by PyTorch-based neural networks for intent recognition, ListenJs for voice input, and Selenium for web automation, while EasyOCR enhances its ability to interact with graphical elements. The system's ability to engage in natural conversations, execute Python code dynam- ically, and correct errors iteratively demonstrates its robustness and versatility. Future improvements in GUI automation, NLP contextual understanding, and deep learning integration will further enhance its capabilities, positioning Jarvis as a valuable tool in various sectors such as public safety, education, and productivity.