Stress is a prominent topic in public health and has permeated all aspects of our everyday lives. Stress has recently become a necessary component of working in the corporate world, particularly in the extremely competitive economy of today. Because of the ongoing mental and physical strain of their jobs, employees in the IT business are more likely to have a variety of health issues. Stress either causes, prolongs, or aggravates diseases. To prevent long-term health problems, it is crucial to identify stress early on and provide medicine. With the use of machine learning and image processing, this research creates non-intrusive techniques for computer-aided diagnostics to identify stress in IT workers. Machine learning models are used to assess visual data taken from continuous video surveillance in various environments, such as work, home, travel, etc., to identify different levels of stress. The creation of machine learning models to assess and identify distinct classes of stress based on multimodal text and visual feature extraction is the main emphasis of the project.