Droplet microfluidics is the field of study which consists of producing and manipulating droplets inside very small channels for different applications that range from particle synthesis to DNA sequencing to soft computing. An experimentalist who fabricates a microchannel and observes the motion of droplets in the channel typically solves the forward problem which involves understanding the behaviour of droplets for a given geometry and operating condition. However, a design problem is an inverse problem as it involves identifying the geometry and operating strategy required to yield the desired behaviour of droplets in a microchannel. And the non-linear nature of droplet flow which results in collective droplet motion renders design non-intuitive. Hence, there is a need for a systematic approach to design which would involve posing the design problem as an optimization problem. However, the multi-scale nature of the flow problem makes the conventional computational fluid dynamics approach infeasible for design. And, without a computationally simple modelling strategy for droplet flow, it will be impossible to uncover the full potential of the droplet microfluidics technology. Hence, during my doctoral work, I proposed to address this problem by introducing the agent-based approach to simulate complex droplet flow and behavior in microchannels.