A computer readable medium comprising instructions which when executed by a computer system causes the computer to implement a method for dynamically improving a process flow is provided. For a plurality of cells, which define activities within the process flow and are characterized by a corresponding plurality of productivity data elements, the method triggers a state engine to simulate the process flow, with the state engine configured to manage all operations rules related to the process flow, and determines at least one process improvement. The method further identifies productivity data elements that correspond to the at least one process improvement, and dynamically modifies the identified productivity data elements to improve the process flow.