Misinterpretation of requirements remains a major challenge in software development, often leading to project failures. Agile Model-Driven Engineering promotes iterative modeling and validation to enhance software quality and adaptability. Business Process Model and Notation (BPMN) serves as a widely adopted standard for visually representing system processes, ensuring a shared understanding among stakeholders. However, the complexity of BPMN can introduce structural errors, compromising model accuracy and reliability. This paper presents a formal verification approach leveraging Object Constraint Language (OCL) to enforce BPMN structural correctness, aligning with Agile principles by enabling early error detection and refinement. Natural Language Processing (NLP) techniques are employed to automate the translation of BPMN standard rules into OCL constraints, enhancing model validation. Empirical evaluation demonstrates the effectiveness of this approach in improving BPMN model accuracy, supporting iterative and reliable model-driven development in Agile workflows.