Companies automate their IT service management decision tasks by deploying increasingly intelligentsystems. However, there is a delicate balance: if the designers over-rely on the automatic capabilitiesoffered by the systems, then there is a risk that the end-users will not trust the algorithmically generatedsolution suggestions to be the correct ones for the task. A more balanced design approach that (i)considers human-centered intent and (ii) can provide managerial decision support that complementshuman capabilities in IT service management is thus needed. We define this as a Hybrid DecisionSupport System (HDSS). HDSSs are a class of decision support systems (DSSs) that extend the class oftraditional DSS to utilize the complementary capabilities of humans and machines for better solutionsto decision tasks. This study uses the action design research method to build, intervene, and evaluatethe designed HDSSs. Five design principles concerning how HDSSs should be built are presented.