All projects// 04
Koalingo
We built an English learning platform that personalises vocabulary with AI: the Oxford 3000 and 5000 lists, a word world organised by level, speaking and writing practice, an exam mode, and the panel that feeds all of it.
Sector
Education Technology / Language
Service
Software, UI/UX, Artificial Intelligence, SEO & GEO
Duration
Ongoing
Year
2025 - 2026
Stack
- React
- TypeScript
- Vite
- Tailwind CSS
- Supabase
- PostgreSQL
- OpenAI
- Technical SEO
- GEO
The Challenge
Most vocabulary apps hand everyone the same list, so learners revise what they already know and never reach what they do not. The platform had to find the real gap, bring it back at the right interval and keep practice from becoming a chore, while also capturing search traffic from thousands of word queries.
The Approach
- // 01A word world was built: the Oxford 3000 and 5000 lists across levels A1 to C1, themes and collections, each word with its meaning, pronunciation, examples and synonyms on its own page.
- // 02Learning state is kept per word, so the study list is rebuilt per person from what is learned, due for revision or never seen.
- // 03AI assisted practice modes were developed: spoken practice, a writing project and a placement test, all fed by the learner's own history.
- // 04An SEO architecture generates a page per word with a word level sitemap, and a GEO layer makes the definitions citable in AI search.
- // 05A content network of guides and blog posts was woven into the word pages through internal links.
- // 06A management panel handles users, word lists, content and subscriptions in one place.
The Results
0
Oxford word list
0
Levels, A1 to C1
0
Practice modes
"Learners work on the words they personally do not know, instead of one list handed to everyone."