With the rapid acceleration of information processing and the increasing volume of digital content, there is a growing demand for intelligent processes across various fields. When dealing with complex business tasks, the traditional static task assignment approach proves to be ineffective and inefficient. We investigate task allocation method of Flowable (an open-source software) and highlight the limitations of its static binding strategy in terms of adaptability and rationality. Based on this analysis, a comprehensive multi-criteria task allocation strategy is proposed, considering factors such as task type, urgency, employee workload, capacity, availability, and workload balance. The multi-objective optimization problem is solved using Hybrid Particle Swarm and Gray Wolf Optimization (HPSGWO), and a multi-dimensional evaluation task allocation model is constructed. In conjunction with the time-constrained push model, an optional time interval is set, allowing employees to proactively choose tasks. If the deadline is reached and the task list remains unallocated, the system automatically reassigns suitable employees based on the multi-criteria strategy. Flowable’s Application Programming Interface (API) is extended to enhance flow control, enabling dynamic routing, task deletion, and re-triggering functionalities. Finally, the optimized Flowable workflow is applied to the management system. The improved Flowable not only enhances user work efficiency and relieves the system from repetitive and tedious task allocation, but also provides more time for complex and intricate business processing.