Today's systems distract users in many ways, thereby reducing their effectiveness. Aura aims to minimize user distractions by trading plentiful computing resources for this scarcest of resources. Aura?s approach is based on two broad concepts. First, it uses proactivity, so a system layer can anticipate requests from a higher layer. Second, it is self-tuning, so layers can adapt by observing current demands and adjusting their performance and resource usage characteristics accordingly. By applying these concepts at every layer in the system, the authors hope to build an environment that adapts to the user?s context and needs with minimal user involvement. They will deploy a large-scale Aura system at Carnegie Mellon to help evaluate the research in a realistic setting.