Scalable Exam File Processor

I’m building a web platform that lets hundreds of thousands of students upload answer-sheet images during exam season, have those files cleaned up, compressed, converted to PDF, and stored safely for evaluators. The front end is already planned in React.js with Next.js for SSR and routing, so you’ll work with both frameworks to craft a snappy, fully responsive interface that handles drag-and-drop and multi-file uploads without hiccups.

On the server side we’ll stay in the Node.js world, using Express to expose REST APIs for every step of the workflow: upload, image manipulation, PDF merge, download, and audit logging. I’m open to either MongoDB or PostgreSQL—choose the one that best fits the schema you design for exam rules, metadata and access logs.

For image processing the stack is definitive: Sharp for fast resizing and format conversion, OpenCV for any background-removal logic that remove.bg can’t cover, and ImageMagick for edge-case transformations. Typical actions include:
• resize, crop, rotate
• convert JPG ↔ PNG ↔ PDF
• compress to a target kilobyte size
• merge multiple PDFs into one booklet

Every file operation must run in a secure, sandboxed flow with checksum validation so no corrupted or malicious content slips through.

The finished app should deploy cleanly to AWS (preferred for scale) but must also run on Vercel or Netlify for staging. Auto-scaling rules, CDN usage and caching layers need to be tuned to survive traffic spikes running into lakhs of concurrent users.

Deliverables I will test against:
1. Source code (frontend, backend, infra scripts)
2. One-click deployment docs or scripts
3. Sample load-test results showing stable throughput at scale
4. A short README explaining how to add new processing rules

If this is squarely within your full-stack and image-processing wheelhouse, let’s talk timelines and milestones.