Six Fixes and One Honest Half-Measure (AICademy Part 2)
A rebuild log for AICademy: the six upgrades I shipped after a code review — scroll restoration, empty-state removal, lazy loading, a week's-path checklist, catalog wayfinding and worked examples — the two findings I declined, and the performance fix I deliberately left half-done.
Part 1 described AICademy as it stands — six courses, 82 modules, simulated employers, a module-scoped AI tutor. What it didn’t mention is that once the thing worked, I put it through a proper code review. The review came back with eight findings.
I shipped six, declined two, and finished one of them only halfway — on purpose.
The six that shipped were mostly small, and that’s the point. Scroll position now resets when you navigate, instead of dumping you halfway down a new page. The Notes tab no longer appears on the 81 of 82 modules that had no notes behind it — an empty state repeated 81 times isn’t a gap, it’s a message. Page code now lazy loads instead of arriving in one 1.28 MB parcel. The module page gained a numbered “This week’s path” strip so seven equal tabs became an ordered route. The catalog gained a role selector that recommends a path and a sequence rather than leaving you to pick blind from eight cards. And every course now ships a worked example — because being graded against a rubric you’ve never seen a finished version of is a quietly unfair ask.
The two I declined are the more interesting half. One asked me to bring the flagship RAG course to feature parity with the newer courses; checking the code showed it wasn’t broken, just built to a different shape before those features existed. The other asked me to tone down the steampunk catalog for an executive audience — a real argument, but that identity is positioning, not a defect. This article is themed after it.
And the performance work is deliberately incomplete. I shipped the safe half; the remaining ~1.1 MB of course data needs a refactor of the core data path, and I couldn’t run a production build to verify it. Shipping a risky refactor you can’t build-test isn’t confidence — it’s gambling with a live site.
Each section pairs the change with an explainer of the technology involved — what a bundle is and why Vite splits it, what wayfinding borrows from airport signage, what “refactor” actually means and why it’s more dangerous than it sounds. No code required.
The academy itself is at ragentic.netlify.app.
About the Author
Ajay Walia
AI {IT Architect} focusing on local-first multi-agent AI engineering, zero-data-egress systems. Ideator, Creator and Executor on Curious Bit.
Keep Reading
An AI Academy for the People Who Keep the Lights On (Part 1)
AICademy: six hands-on AI courses and 82 modules built for infrastructure support engineers at GSIs like TCS, HCL and Infosys. The philosophy, the stack, the courses, and why it's different from every AI course online today.
Choose Your Path — Five AI-Augmented Careers, One Academy
A visual guide to AICademy's five hands-on paths — Service Desk, SysAdmin, IT Project Manager, RAG Engineering, and Agentic AI Leadership — and how to pick the one that matches the job you already do.
Building a Support Agent That Knows When to Stop — My Agentic AI Capstone
A technical deep-dive into my IIT Madras Agentic AI capstone: a framework-free customer-support resolution agent with deterministic pre-LLM safety, grounded RAG, guarded tool use, memory — and the live date-arithmetic failure that reshaped its architecture.