I built an anonymous teacher review site for my university. The code was the easy part.

Rajshahi University of Engineering & Technology student constructed an anonymous faculty review platform called RUET Faculty Review. This platform aims to collect unbiased student feedback on teaching quality without revealing the identity of the reviewers. The site uses cutting-edge technologies like Next.js, Prisma, Postgres, and Vercel to build its backend and frontend.

The creator of RUET Faculty Review encountered three significant challenges during the development process. First, the moderation system proved to be problematic. Initially, the plan was to have users submit reviews, which would then be reviewed by an admin. However, the admin never opened the queue, as students would be left hanging with unread reviews during exam periods.

To solve this issue, the platform now utilizes an AI moderation system, which classifies reviews as pending. The AI's inaccessibility has led to low moderation activity, necessitating a retry system that alerts the user only in case of prolonged unavailability.

Secondly, the platform initially had overly strict filters that rejected honest criticism. For instance, a review claiming a teacher explained concepts too quickly and skipped questions was rejected. The solution was to adopt a "DEFAULT TO SAFE" approach, where harsh, angry, or informal criticism is allowed, whereas personal attacks, accusations, or spam are not. This change has ensured that the site collects genuine and constructive feedback from students.

Thirdly, the platform faced an issue with user authentication. A helper function, teacherCapProblem(userId, teacherId), was exported from a use server file, enabling users to check if a particular user had reviewed a specific teacher. However, this function was accessible even to unauthenticated users, which contradicted the site's core principle of anonymity. To resolve this issue, the helper was moved to a plain module, eliminating the endpoint and preventing unauthorized access.

In conclusion, building a successful anonymous review platform like RUET Faculty Review requires careful consideration of various aspects, such as moderation systems, filtering strategies, and user authentication. By addressing these challenges, the creator managed to build a platform that encourages honest and constructive feedback on teaching quality, without revealing the identity of the reviewers.