During exam week, thousands of college students open notes simultaneously from campus hostels with notoriously congested Wi-Fi. Forcing students to download raw 60MB to 120MB PDF scans was unacceptable.
The On-Demand Page Streaming Solution
We built an in-browser PDF rendering pipeline powered by PDF.js web workers and HTML5 canvas. Instead of downloading the full document upfront, we stream only the pages currently visible in the student's viewport.
Initial load times dropped from 22 seconds down to 350 milliseconds. Mobile battery consumption plummeted, and memory leaks on budget Android devices disappeared.
“Good university software must be engineered for real-world campus constraints—spotty 4G, low-end mobile hardware, and high exam-night stress.”
Adding Smart Syllabus Bookmarks
We parsed notes into unit-wise chapters so students can jump straight to 'Unit 3: Normalization & BCNF' with one tap without scrolling through 150 pages.
Engineering Deep Dive: Web Worker Architecture
Complete Next.js implementation snippets, custom canvas virtualizers, and edge CDN cache header configurations.
How we eliminated browser lag when rendering complex circuit diagrams and handwritten equations.
