The paper proposes location/application specific tuning of the solar power generation curve by appropriately orienting solar panels in a solar tree. The initial part of the study involving optimization of single panel orientation, emphasizes the need to adopt data driven approach. Study carried out for 15 locations covering a large latitude range shows that optimal orientation cannot be determined based on the latitude angle alone and azimuth angle also needs to be considered, especially for locations having asymmetric solar insolation pattern during a day. Based on these insights, feasibility of tuning annual solar power generation curve is shown by designing a 1 kW solar tree for four locations. Genetic algorithm based optimization is used for positioning of the solar panels so that shadow losses are minimized. Validation with ray optic simulations for solar trees designed for two locations have shown less than 2% shading losses. The optimized solar trees for San Francisco and. Paris show increase in power generation of 2.04% and 7.38% respectively compared to latitude tilt. The simulations are further validated using a scaled down prototype solar tree showing excellent match. The methodology presented here can be easily extended to the design of optimized solar tree for any location and capacity. (C) 2018 Elsevier Ltd. All rights reserved.