Context
A globally recognised university in Western Australia known for its strong industry ties, high-impact research, and use of cutting-edge tech worked with Graide to evaluate the impact of AI-assisted marking on efficiency, accuracy, and feedback quality. The pilot primarily involved 100 essay submissions to determine if Graide could reduce marking time, improve feedback quality, and scale high volumes. Graide shows compelling promise as a scalable solution for modern assessment needs in higher education.
The Challenges
A lot of the grading process involved repetitive administrative steps, such as applying similar comments to different students, which took time away from higher-value feedback.
As student numbers grow, the university needs a system that could scale up efficiently without increasing staff burnout or compromising quality.
Marking long, essay-style assessments took a long time, making the process slow and resource-intensive, especially at scale.
Results
Marking got faster and smarter. On average, it took just under 18 minutes to grade each script — that’s a 17% time saving compared to the usual 20 minutes. Even better, the median time dropped to 16.6 minutes
Students got detailed, high-quality feedback, and markers didn’t have to repeat themselves endlessly.
75% of the grades suggested by the AI matched exactly what the human marker chose. This indicates that the AI consistently provided accurate results.
