Workplace activity recognition (WAR) is a subset of human activity recognition (HAR) which is used to automatically recognize specific activities in the workplace. Conventional deep learning-based approaches lack context-awareness to be deployed in diverse workplaces and industries for a generalized WAR model. These models are restricted to the designed workspace, thus highlighting a major obstacle in both WAR and HAR, which is the lack of relevant data. A general WAR model provides the foundations needed for general workplace safety and prevention. A context-aware combination that leverages the pattern recognition strengths of a deep learning model and the contextual understanding of a large language model can potentially solve this issue, removing the need for specialized data for every workplace application by limiting the data required to just bodily motion and discerning the likely activity from the contextual data available.