Solar photovoltaic (PV) systems have a utilization (or capacity) factor of 15–20% worldwide. We propose to enhance the energy yield in a software-defined manner by complementing commodity solar PV systems with cloud-based IoT-controlled reflectors. We also propose designs for brownfield and greenfield settings in solar farms. We study a number of practical engineering issues including effect of solar azimuth, shadowing effects, ground coverage ratio (GCR) tradeoff, constraints on angular control etc. Our designs can raise solar PV energy yield between 50–100% with modest tradeoffs on operational complexity, land requirements (ground coverage ratio) etc. The software-defined IoT control allows a variety of current and future operational or business constraints to be flexibly factored in to tradeoff these factors versus economic gain (eg: levelized cost of energy, LCOE). The paper presents both simulation and experimental evidence for our system. We are actively piloting this technology with solar PV developers and engineering, procurement, construction (EPC) companies in emerging markets.