The identification of bioactive natural products (NPs) in complex mixtures has become an important subject of contemporary NP research. In an attempt to address this challenge, the present work proposes an integrated strategy that combines tandem mass spectrometry (MS2)-based molecular networking (MN), a partial least-squares (PLS) chemometric model, as well as 13C NMR-based dereplication using MixONat software. In addition, an advanced glycation end product (AGEs) assay was used for activity evaluation. The approach was implemented on a Garcinia parvifolia bark extract that comprised a high content of prenylated xanthones and had previously shown a notable inhibitory effect on AGE formation. As a main result, the proposed strategy permitted the identification of potentially active metabolites within complex mixtures and their annotation with a higher level of confidence by NMR data. Overall, this comprehensive approach provides a powerful and efficient solution for the targeting and annotating of active compounds in complex NP mixtures.