Innovative Multi-Language Sentence Analysis Tool Reaches Nearly 1,000 Users
In February of this year, I embarked on an exciting journey to develop a multi-language sentence analysis platform for my wife, starting with a simple, single-page MVP built with Next.js. The core idea was to utilize the Gemini API to generate color-coded Korean grammar insights, providing a helpful learning aid. What began as a proof of concept quickly evolved into a robust online tool called Hanbok, which now processes hundreds of analyses daily—totaling over 40,000 instances—and has attracted nearly 1,000 dedicated users.
Throughout its development, Hanbok has undergone numerous redesigns and feature enhancements. Notable additions include spaced repetition flashcards, support for nine additional languages, song lyric analysis, and a conversational chat module that allows users to ask follow-up questions based on specific sentence analyses. These features aim to provide a comprehensive language learning experience tailored to user needs.
You can try out the platform here: https://hanbokstudy.com — feedback is welcome!
Technical Stack Overview:
– Frontend: Next.js with custom CSS styling
– Backend: Express server
– Data Storage: Redis and MongoDB
– AI & OCR: GPT-4.1 integrated for OCR and conversational interactions
– Analysis Engine: Gemini API
– Deployment: DigitalOcean with NGINX
Looking ahead, I plan to introduce a community forum dedicated to language Q&A, expand language support based on user requests—such as Vietnamese, Hindi, and Indonesian—and develop grammar practice quizzes tailored to individual learning histories.
For those interested in the development process or community engagement, the project’s GitHub repository and Discord server are linked on the site. Your support and feedback are greatly appreciated as this project continues to grow.
