Exam Plagiarism Detection Integration

I am enhancing our online proctoring platform with automated cheating detection, and the immediate priority is reliable plagiarism detection. The goal is to plug a best-in-class plagiarism engine into our existing exam workflow so that every student submission—whether typed into a web-based essay box or uploaded as a file—is scanned in real time before it is released for grading.

Here is what I need from you:
• Recommend or confirm a robust plagiarism-detection API (Turnitin, Copyleaks, or an open-source alternative if it meets accuracy requirements).
• Map out the integration flow: trigger points, data sent, callback/webhook handling, and how results are surfaced to instructors inside our dashboard.
• Build and document the code that connects our backend (Node + PostgreSQL) to the chosen service, including retry logic and secure storage of similarity reports.
• Create a lightweight admin panel page that shows match percentages, links to sources, and a plain-language risk flag so instructors can act quickly.
• Write unit tests and provide a brief deployment guide so our internal team can maintain the module.

Although plagiarism detection is the only mandatory feature right now, I am keeping the door open for future additions such as AI-based behaviour analysis or audio anomaly checks, so please structure the solution in a modular way.

Acceptance criteria: end-to-end demo on our staging server showing a submitted essay, an automatic plagiarism scan, and a report displayed inside the instructor view within 3 seconds on average. Provide source code, API keys redacted, and clear setup instructions.