The invention uses advanced computer science techniques and sophisticated machine learning models to help predict how employees will perform. Employee messages, performance evaluations, project records and feedback are among the data types used by the system to apply GPT-4 and BERT models to get fine details about each employee’s behavior, abilities and performance. The technique analyzes emotions, gathers topics and processes words in context to determine both open and hidden measures of performance. Our model uses NLP, employee history and organizational data to provide reliable, on-demand evaluations of each employee. Through this method, managers can manage talent actively, design personal learning plans, build better teams and make informed choices. Because it is objective, continues throughout the year and can handle large numbers of employees, this invention improves workforce results and keeps employees in companies when there are big changes in the business.