Enzymatic pretreatment plays a crucial role in producing cellulose nanofibers (CNFs) before fibrillation. While previous studies have explored how treatment severity affects CNF characteristics, there remains a lack of suitable parameters to monitor real-time enzymatic processes and fully comprehend the link between enzymatic action, fibrillation, and CNF properties. This study focuses on evaluating the impact of enzyme charge (using a monocomponent endoglucanase) and treatment time on cellulose fiber morphology and reducing sugar generation. For the first time, a random forest (RF) model is developed to predict reducing sugar concentration based on easily measurable process conditions (e.g., stirrer power consumption) and fiber/suspension characteristics like fines content and apparent viscosity. Polarized light optical microscopy was found to be a suitable technique to evaluate the morphological changes that fibers experience during enzymatic pretreatment. The research also revealed that endoglucanases initially induce surface fibrillation, releasing fine fibers into the suspension, followed by fiber swelling and shortening. Furthermore, the effect of enzymatic pretreatment on resulting CNF characteristics was studied at two fibrillation intensities, indicating that a high enzyme charge and short treatment times (e.g., 90 min) are sufficient to produce CNFs with a nanofibrillation yield of 19–23 % and a cationic demand ranging from 220 to 275 μeq/g. This work introduces a well-modeled enzymatic pretreatment process, unlocking its potential and reducing uncertainties for future upscaling endeavors.