This paper proposes a conceptual approach for interactive process discovery and improvement in people-driven business process. Developed as a supplementary element to traditional process mining, it addresses two recently identified challenges of process mining highlighted in the Process Mining Manifesto by the IEEE Task Force on Process Mining. The developed approach guides users through the execution of business processes via recommendations based on underlying process models, which can result from traditional process mining techniques. Specifically, the approach provides recommendations learned from previous processes executed by a particular user and couples them with decisions taken from all users involved in that business process. The individual process executions eventually feed back into the overall process model enabling process model evolutions and hence living process models.