😭 The Problem
This project has a slightly unhinged origin story.
Years ago, I was learning Japanese and had a Japanese friend who would tease me because, despite studying the language, I could not read very well.
And... fair.
I could recognize some Japanese. I knew vocabulary. I could understand things in context. But recognizing a kanji when it was sitting in front of me was very different from actually being able to read it, remember it, and write it myself.
Then I found an Android app called KanjiTree.
KanjiTree was simple, inexpensive, and incredibly useful to me. It connected kanji recognition with vocabulary, readings, and — most importantly — handwriting practice.
Instead of only asking:
"Do you recognize this character?"
it made me actually interact with Japanese writing.
And it helped.
For the first time, I felt like I was developing Japanese literacy, rather than getting better at guessing answers on language-learning apps.
Eventually, I stopped using it.
Years passed.
Then KanjiTree disappeared from the app store.
I assumed it was just another abandoned indie app.
Fast-forward to 2026.
I'm on my way to Atlanta for RenderATL, thinking about what I might build for a Major League Hacking project. I suddenly remembered that weird little Japanese app that had helped me so much.
So I searched for KanjiTree.
And that is how, on the ride to Atlanta, I discovered that the reason my favorite kanji app disappeared was apparently much darker than a deprecated Android API:
its developer had gone to prison for murder.
That was... not the product lifecycle update I was expecting.
After processing possibly the wildest piece of software-maintenance lore I've ever encountered, I kept thinking about the app itself.
KanjiTree had been remarkably simple.
It didn't need a giant subscription.
It didn't need to turn learning Japanese into a social network.
It didn't need generative AI.
I remember paying about $9.99 for lifetime access, and that little app did something genuinely valuable for me: it helped me learn to read and write Japanese.
So I decided:
The idea deserves to survive even if the original app didn't.
That's why I started building Kaku.
🌸 The Solution
Kaku is an affordable Japanese literacy trainer built around a simple principle:
Recognition alone is not literacy.
Knowing that 山 means "mountain" when you see it doesn't necessarily mean you can read it in a word.
And being able to read it doesn't necessarily mean you could pick up a pen and write 山 from memory.
Kaku therefore treats Japanese literacy as several connected skills:
Recognize → Read → Write → Review
Instead of giving learners one generic "mastery" score, Kaku is designed to track those abilities separately.
A learner might eventually see:
山
Recognition ██████████ 100%
Reading ███████░░░ 70%
Writing █████░░░░░ 50%
Built With
- codex
- render

Log in or sign up for Devpost to join the conversation.