Organizations operating in highly regulated environments rely on information systems to implement and enforce process rules that ensure consistent and compliant process execution. Evaluating whether these rules are faithfully reflected in system behavior is therefore essential for both process governance and operational reliability. This study employs conformance checking to evaluate the implementation of process rules in the loan approval process of a major Iranian bank. A structured five-phase process mining methodology was applied to event logs comprising 9,966 recorded activities across 774 loan cases. The extracted event logs were cleaned and compared against a reference process model constructed from official documentation and expert knowledge. The analysis showed that although the reference model permitted only three valid execution variants, the event logs contained 24 observed variants. Further investigation revealed that many apparent deviations did not represent actual process rule violations. Instead, they primarily resulted from delayed activity registration, incomplete timestamps, and manual recording practices, which caused discrepancies between the executed process and its representation in the information system. Temporal analysis further indicated that 31.3% of cases exceeded the prescribed processing time. The findings demonstrate that conformance checking can support not only compliance assessment but also the evaluation of how process rules are implemented within information systems. Based on the identified shortcomings, the study proposes system-level improvements, including workflow constraints that prevent progression to subsequent activities before mandatory task completion is recorded. These recommendations contribute to improving process traceability, enhancing the reliability of event data, and strengthening process governance.