Ask a room of CS students what a red-black tree rebalance looks like, and you'll get silence. Not because they're not smart — because nobody ever let them see it. Data structures are taught as static diagrams for something that is fundamentally a thing in motion.
This is the gap I got obsessed with, and it's why I built AlgoVerse — a platform where you don't read about algorithms, you watch them think.
The real reason DSA is hard
An algorithm isn't a picture. It's a sequence of decisions unfolding over time — a pivot chosen, a comparison made, a pointer moved. A textbook flattens all of that into one frozen snapshot and a Big-O label, and then we act surprised when students can memorize "quicksort is O(n log n)" without any feel for why it collapses on a sorted array.
You can't build intuition for something you've never seen move. That's not a talent problem. It's a tooling problem.
What I built
AlgoVerse turns the abstract mechanics of data structures and algorithms into animated, explorable 3D scenes you can start, pause, and step through at your own pace:
- Sorting, live — watch comparisons and swaps happen in real time, and feel the difference between an O(n²) sort and an O(n log n) one instead of being told about it.
- Trees that rebalance in front of you — insertions, rotations, traversals, animated as they occur, so an AVL rebalance stops being a page of arrows and becomes something you've actually watched.
- Graphs you can build and run — place nodes, add edges, and watch BFS, DFS, and shortest-path algorithms explore.
- Complexity you can interrogate — drag the input size up and watch the growth curves diverge, so the cost an algorithm "has" becomes a thing you observe, not a phrase you repeat.
It's built on React 19 and Three.js, and the genuinely hard engineering wasn't the algorithms — those are well understood. It was decoupling the computation from the rendering so large inputs never freeze the UI, and choreographing the motion so it teaches instead of just decorating.
Why it matters
Educational tooling has a reputation for looking like it was built in 2009, and mostly it earned that. I wanted to prove learning tools can carry the same production polish as any consumer app — not because polish is the point, but because it removes the friction between a learner and the moment a concept clicks.
The moment something clicks is almost always the moment you see it. AlgoVerse is 12 modules built entirely around engineering that moment.
If you've ever stared at pseudocode wishing you could just watch the thing run, that's exactly who I built it for. The full breakdown, the modules, and the live demo are on the project page.
👉 See it: www.divyakush.com/projects/algoverse
Divyakush Punjabi — Full-Stack & AI Systems Engineer
🌐 https://www.divyakush.com · 💼 LinkedIn · 💻 GitHub
