Available to hire
Hi, I’m Yongjae Kim, a frontend engineer from Seoul, Korea. I have four years of frontend experience in production SaaS and I’m currently building full-stack AI products. I enjoy turning complex problems into intuitive, scalable interfaces and systems.
I built and operate Nativ, a PDF-based AI chat tutor for language learners, with a hybrid RAG pipeline and LLM-as-Judge evaluation in production. I’m passionate about cost-efficient architectures, thorough testing, and shipping end-to-end features—from UX design to deployment.
Check out my portfolio for more details: Website not available. Sign in: https://www.twine.net/signup
Skills
Experience Level
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Expert
Intermediate
Language
Korean
Fluent
English
Advanced
Work Experience
Frontend Engineer at VNTG
November 18, 2025 - February 11, 2026Digital transformation platform for manufacturing enterprises. Defined a roadmap to automate frontend code generation from Figma designs and requirements. Built a 4-stage Spec-to-React pipeline (Claude Code sub-agents) with automated quality gates and auto-correction (WCAG AA, semantic HTML), improving HTML quality from 50–70% to 95% and achieving 90%+ Figma-to-React component mapping accuracy.
Frontend Engineer at Bold9 Ltd
June 7, 2021 - June 12, 2025SaaS e-commerce fulfilment platform for online sellers. Enhanced core business pages with advanced search, interactive data tables, and inline editing; created a real-time operations dashboard; developed large-scale forms (70+ fields) with dynamic validation to reduce re-renders; built a cross-platform push notification system; introduced automated unit, integration, and visual testing with Vitest and Storybook, increasing test coverage from 10% to 70% and cutting CI runtime by 50% via parallel tests.
Founder & Full Stack Engineer at Nativ
February 25, 2026 - PresentBuilt and operating a live AI tutor for language learners, solo. Designed hybrid RAG pipeline (pgvector + BM25, fused with RRF), cross-model LLM-as-Judge eval catching quality regressions on every prompt change, and 163 backend integration tests in CI/CD. Launched as freemium SaaS with subscription billing and Docker deployment on $3–8/month infrastructure.
Overseas Sales Executive at DRGEM Corporation
January 22, 2018 - February 29, 2020Built partnerships with Fujifilm branches across Asia, ME, and Africa. Attended global medical exhibitions (RSNA, ECR, ARAB HEALTH) to gather market insights and influence product strategy.
Frontend Engineer at Bold9 Ltd
February 1, 2026 - PresentSaaS e-commerce fulfilment platform. Enhanced core pages with advanced search filters, interactive data tables with inline editing, bulk operations, and multi-column sorting at smooth 60fps. Built a real-time operations dashboard with visual widgets for orders, market data, and invoices. Developed large-scale forms (70+ fields) using Formik/Yup with dynamic validation to reduce re-renders. Implemented a cross-platform push notification system and automated testing with Vitest/Storybook, increasing test coverage to 70% and cutting CI test runtimes by 50% via parallel execution.
Education
Bachelor of Computer Science at Korea National Open University
March 1, 2022 - August 21, 2024Bachelor of Economics at Soongsil University
March 1, 2010 - February 9, 2018Full-stack course based on JavaScript at Codestates
October 15, 2020 - March 13, 2021Bachelor of Computer Science at Korea National Open University
January 11, 2030 - April 25, 2026Full-stack course based on JavaScript at Codestates
October 1, 2020 - March 1, 2021Bachelor of Economics at Soongsil University
March 1, 2010 - February 28, 2018Qualifications
Industry Experience
Software & Internet, Manufacturing, Professional Services, Education, Media & Entertainment
- Fine-tuning was never an option — every user uploads their own textbook. Designed RAG with hybrid search (semantic + keyword) so the AI always answers from context, not generic knowledge.
- Language tutoring needs consistent accuracy over reasoning depth — switched from Anthropic to OpenAI and right-sized models per task — A/B tested for quality, cost modeled at 94% reduction ($133 → $8.60/month) at scale.
- 163 backend integration tests in CI/CD, plus a cross-model LLM-as-Judge eval (GPT tutor → Claude judge) that automatically catches RAG quality regressions on every prompt change.
- Designed a text-selection popup inside the PDF viewer — translate, speak, practice pronunciation, save to vocabulary, or ask the AI tutor — so learners stay focused without ever leaving the page.
- Launched as a live freemium SaaS with subscription billing and Docker deployment on $3–8/month infrastructure.
PDF-based AI chat tutor for language learners
Hire a Full Stack Developer
We have the best full stack developer experts on Twine. Hire a full stack developer in Seoul today.