Continuous Integration (CI) is a hot topic. Yet, little attention is payed to how CI systems work and what impacts their behavior. In parallel, bug prediction in software is gaining high attention. But this is done mostly in the context of software engineering, and the relation to the realm of CI and CI systems engineering has not been established yet. In this paper we describe how Project Gating systems operate, which are a specific type of CI systems used to keep the mainline of development always clean. We propose and evaluate three heuristics for improving Gating performance and demonstrate their trade-offs. The third heuristic, which leverages state-of-the-art bug prediction achieves the best performance across the entire spectrum of workload conditions.