Labor shortages across industries are increasing the demand for robots and skilled robot programmers. Learning from Demonstration (LfD) offers a promising approach to address this challenge, allowing robots to acquire skills by observing human demonstrations. Existing handheld devices for LfD are often task-specific and limited to particular learning phases. This work presents a modular handheld data acquisition device for LfD, designed to support multiple tasks and learning methods with a single platform. The device features a robotic flange to enable quick swapping of end-effectors and their associated sensors, providing flexible data collection across different tasks. An ArUco marker cube facilitates basic pose and position tracking, while ROS2 integration supports data communication and system control. Validation experiments demonstrated reliable performance, and user evaluations confirmed high ease of use and comfort. By offering a versatile and user-friendly platform, this modular device reduces the need for multiple task-specific handheld devices and broadens its applicability in LfD research and practical applications. Future work will focus on mapping the acquired demonstrations onto specific robots while accounting for kinematic constraints and workspace limitations, enabling seamless task reproduction.