This article introduces UserPrint, a behavioral fingerprinting approach that characterizes corporate users based on their software usage patterns. A real dataset comprising 41,859 users collected over six months is analyzed to evaluate the temporal consistency of application usage through cosine similarity and cosine distance. The representation is further extended with web navigation signals using a controlled synthetic dataset derived from a real user profile, enabling the assessment of how additional modalities affect profile separability. The results demonstrate that cosine-based metrics reliably capture both stable behaviors and meaningful temporal changes, while the inclusion of navigation data substantially increases user separability. These findings indicate that continuous behavioral metrics and multimodal feature representations provide a promising foundation for security monitoring and risk detection in corporate environments.