In this project, we present a concept to detect employee stress using machine learning algorithms such as Support Vector Machine (SVM) and Random Forest. Social media datasets, such as tweets, analyze employee moods and detect stress levels. Used a pre-processed Twitter dataset with the Natural Language Toolkit (NLTK) to remove stop words and special characters, and trained and tested the machine learning models using 80% and 20% of the dataset respectively. The results show a stress detection accuracy of over 90% using the Random Forest and SVM algorithms. The SVM algorithm creates a line or hyperplane that separates the data into classes and the Random Forest algorithm selects random values to classify data points.