The energy sector is in the midst of an exciting transition. Driven by new generation technologies and by infrastructure digitalization, the traditional way of transmitting, distributing and using energy is transforming a centralized hierarchical system into a multi-directional open infrastructure. While the vision of Intelligent Energy Networks is appealing and desirable---especially from a sustainability perspective---a number of major challenges remain to be tackled. The loss of centralized control, the intermittent nature of renewable energy sources and the scale of the future digital energy systems are novel situations for power systems infrastructures and consumers that pose reliability and availability threats. In this talk, I show examples of how Computer Science techniques are having and will have an important role in future energy systems. I focus on electricity as energy vector, and techniques from Service-Oriented Computing and AI Planning. I also present Complex Network theory as a design tool for energy distribution systems. To make things concrete, I will review almost ten years of personal research that include making office buildings energy efficient, homes smarter, and futuristic models for the evolution of power distribution grids to accommodate for multi-directional energy flows with distributed generation and local control.