Declarative process mining has emerged as a powerful para-digm for modelling, analysing, and improving knowledge-intensive processes, offering flexibility and expressiveness beyond traditional procedural approaches. In this paper, we present a comprehensive overview of recent advances in declarative process discovery, with particular emphasis on Dynamic Condition Response (DCR) Graphs and the DisCoveR algorithm. We present a high-level overview of the base algorithm, discuss important extensions and successors that allow the mining of hierarchy, the use of labelled event logs, the mining of object-centric models, and the mining of probabilistic models, and provide a short overview of important enabling technologies such as a mapping to Petri nets and algorithms for model-log alignment. Finally, we briefly outline a number of key open questions and future research directions.