Abstract: Energetic jets are rare auto-generated probes that offer the opportunity to reveal the inner workings of the most ideal liquid in Nature, the quark-gluon plasma (QGP). This new phase of deconfined QCD matter, which filled the Universe a few microseconds after the Big Bang, is today produced in heavy-ion collisions at RHIC and the LHC. As jets traverse the flowing QGP, they get modified via color interactions, and modify the medium in turn. However, due to the multi-scale and multi-partonic nature of jets, extracting precise information about the properties of the QGP from jet observables is a challenging task. In this talk I will present newly developed strategies aimed to disentangle between different competing phenomena, ranging from rapidity scans of jet substructure observables to machine learning techniques that quantify the degree of jet modification.