Consumers with distributed energy resources (DERs) and intelligent building energy management systems (BEMS) can achieve significant electricity bill reduction by participating in demand response (DR) programs. DR encourages the active consumer to alter their demand profile in response to time-varying electricity tariff or economic incentives. However, large-scale adoption of such DR schemes and BEMS in neighborhood area can lead to rebound peaks when electricity is cheaper. This paper proposes a unified approach to energy management system for a residential neighborhood that mitigates the demand peaks of the neighborhood and provides electricity bill reduction to the participating consumers. An optimization model has been developed that minimizes the energy cost of the participating consumers while reducing aggregated demand peaks in the neighborhood by scheduling the charge-discharge operation of residential storage units and energy sharing among neighbors. The effectiveness of the proposed neighborhood energy management system (NEMS) model is evaluated via a case study for a residential neighborhood in Sydney, Australia with actual electricity meter data and time of use (TOU) tariff structure for the neighborhood. Simulation results indicate the efficacy of the proposed NEMS framework and optimization model in terms of electricity bill reduction and demand peaks mitigation.