The embedded class processors found in commodity palmtop computers continue to become increasingly capable while retaining an energy-efficient footprint. Palmtop computers themselves, including smartphones and tablets, provide a small form factor system integrating wireless communication and non-volatile storage with these energy-efficient processors. In this paper, we present Ocelot, a grid-like computing environment that leverages palmtop computers in place of traditional workstation or server class machines to compute highly parallel light to medium-weight tasks in an energy efficient manner. Ocelot is motivated to be used in a wireless sensor network environment where palmtop computers with onboard or attached sensors form nodes to obtain and process sensor data. Using a centralized server/dashboard, data processing jobs and managed and distributed to these palmtop computers to compute the results in parallel. Our results demonstrate that these palmtop machines provide significant energy advantage - average of 67% in our case study - over energy-optimized desktop, high-performance workstation and server class machines.