Code review remains one of the most important quality-assurance activities in modern software engineering, yet manual review of pull requests is often delayed by reviewer workload, inconsistent practices, and growing repository activity. This paper presents Gen-AI Code Reviewer, a full-stack web platform that automates first-pass pull request analysis using a large language model integrated with GitHub-based development workflows. The system authenticates users through GitHub OAuth, retrieves repository and pull request metadata through the GitHub REST API, extracts changed-file diffs, and submits structured review prompts to an AI model for contextual analysis. The generated output includes a summary, risk score, severity- tagged comments, and suggested fixes, which are persisted in a PostgreSQL database and displayed through a dashboard interface. The platform is implemented using Next.js, TypeScript, Prisma ORM, tRPC, React Query, Better Auth, and Inngest for asynchronous workflow execution. The proposed system is designed to reduce review turnaround time, improve consistency of early feedback, and support human reviewers by surfacing potentially risky changes before manual inspection. The paper describes the motivation, architecture, workflow, implementation strategy, and practical relevance of the prototype, and discusses limitations related to model reliability, repository context, and prompt-length constraints.
Sakhare et al. (Thu,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: