Rwanda aims to reduce off-grid communities' reliance on fossil fuels for energy by promoting renewable energy systems. This study focuses on evaluating and optimising such systems. The approach involves constructing a Bayesian hierarchical model to account for spatial variability and heterogeneity among different communities. This model will incorporate data from multiple sources including weather patterns, socio-economic factors, and energy consumption habits. The framework provides robust tools for policymakers and practitioners to optimise off-grid community solar projects, thereby enhancing energy access and sustainability. Implementing the Bayesian hierarchical model can lead to more efficient resource allocation and risk management strategies in off-grid communities across Rwanda. Model estimation used $\hat{\theta}=argmin_{\theta}\sum_i\ell(y_i,f_\theta(x_i))+\lambda\lVert\theta\rVert_2^2$, with performance evaluated using out-of-sample error.