University examination papers play a crucial role in the institution's quality, impacting the institution's accreditation status. In this context, ensuring the quality of examination papers is paramount. In practice, however, manual assessments are mostly laborious and time-consuming and generally lack consistency. The last decade has seen digital education acquire immense interest in academic discourse, especially when developing intelligent systems for educational assessment. The presented work proposes an automated system that allows text analysis and evaluation of university exam papers by formal and technical criteria. The research was conducted by analyzing 30 exam papers, which will be included in each of the exam papers, which consist of 60 questions each, in total it holds 1,800 questions. Moreover, it also includes research to understand the quality and relationship with students' test anxiety. A total of 50 year one first-year students were taken to measure students' academic stress by a scale. Planning on basic levels and adherence to technical standards were missing in the exam papers. The proposed automated system has improved exam paper quality to a great extent and reduced academic stress among students with an accuracy of 98% in identifying and matching specified criteria.