Mentis : An AI-Powered Platform for Automated Academic Assessment and Rubric-Based Feedback

Academic institutions face increasing challenges in evaluating students’ electronicprojects and assignments, as manual grading requires substantial time and effort and isoften prone to variations in consistency and objectivity, in addition to insufficientlyorganized feedback delivered to students. With the expansion of digital education andthe growing volume of electronic submissions, there is a clear need for technologicalsolutions that support the evaluation process and improve its efficiency withoutcompromising academic standards. In response to this challenge, a supportive webbased system called Mentis is designed to assist the academic assessment process byleveraging predefined evaluation models (rubrics) and artificial intelligencetechniques, ensuring a more organized grading workflow and enhanced efficiencywithout transforming the system into a full learning management platform. Thesystem enables instructors to define evaluation criteria and assign their relativeweights, and to upload students’ submissions in supported textual formats. It thenanalyzes the textual content of these submissions according to the specified criteriaand generates an overall score along with detailed scores for each criterion, inaddition to structured textual feedback that supports academic decision-making.To achieve these goals, the project adopted an Incremental DevelopmentMethodology supported by iterative practices within the Software Development LifeCycle (SDLC). Development was organized into successive functional incrementscovering requirements analysis, system design, implementation, integration, andprogressive testing. This approach resulted in an integrated system that supportsweighted evaluation-criteria management, submission processing, AI-assistedevaluation, structured result presentation, and persistent storage for later retrieval.Mentis is intended to reduce the time and effort associated with grading and improveevaluation consistency by applying clear, instructor-defined criteria while preservinginstructor oversight. Accordingly, Mentis demonstrates a practical and systematicapplication of artificial intelligence to support academic assessment, withopportunities for future extension and integration with broader educational systems.