Persistent atrial fibrillation (PsAF) is associated with high recurrence rates despite pulmonary vein isolation (PVI), highlighting the need to target the atrial substrate. Substrate characterization beyond the pulmonary veins using functional mapping can potentially unmask latent slow conduction areas. To improve substrate characterization in PsAF, we developed an automated tool to quantify key electrogram (EGM) features such as duration, local activation time (LAT), number of deflections, and pacing-induced duration increase (Delta). The clinical data comprise over 1,000 intracardiac EGMs from 5 patients with PsAF, recorded during baseline sinus rhythm and after three short-coupled extra stimuli. The features calculated by the tool achieved intraclass correlation coefficients of 0.98 for R3-LAT and 0.77 for R3-Duration, respectively, when compared to blinded expert clinicians’ annotation, indicating reliable and acceptable automated measurements. Furthermore, the extracted features from the paced EGMs showed strong discriminatory power in identifying hidden slow conduction substrates, achieving 93.6% balanced accuracy with a Support Vector binary classifier model using a linear kernel.