Last update: 2025-10-13

Generalist Frontend Developer

Dohyun Ko

A front-end developer, I have led web and app development in startups and labs, with extensive experience from early development to user analytics. Focused on developing user-centered services utilizing a diverse technology stack. I also run conferences and work to vitalize the local developer ecosystem

Skills

React

TypeScript

Next.js

Python

NestJS

FastAPI

Flutter

C++

Docker

Educations

GIST

Department of Electrical, Electronics, and Computer Engineering

2022. 2 –

TGPA 4.04 / 4.5

Attended Courses

Object-Oriented Programming|Data Structures|Introduction to Algorithms|Computer Systems Theory and Practice|Automata Theory|Operating Systems|Computer Graphics|Computer Networking|Machine Learning and Deep Learning|Computer Security|Computational Models of Cognition

Publications

O2ARC 3.0: A Platform for Solving and Creating ARC Tasks

IJCAI 2024 | Co-first Author - 2024. 08

To address the issue of insufficient human solving data for ARC problems and the poor quality of collected data for AI training, O2ARC 3.0 was developed. This paper introduces the method of improving data collection quality by reorganizing the grid manipulation logic of existing tools and increasing the volume of data collection through gamification.

Currently working in

Channel Corp.

Software Engineer

2025. 01 – Now

ALF Development

2025. 04 – 2026. 06

Developed ALF v2, Channel Corp.'s AI customer support agent.

  • Developed the agent's knowledge base UI

  • Developed the agent workflow UI

  • Reduced costs by 24% by optimizing the key-value (KV) cache

CoS Development

2026. 07 – Now

Developing CoS, an agent that automates customer data integration, analysis, and business tasks.

  • Developed the agent system backend

  • Reduced prompt noise

Experiences

GIST Data Science Lab

Intern

2024. 01 – 2024. 05

O2ARC 3.0

2024. 01 – 2024. 05

Developed O2ARC 3.0 to collect human solving data for ARC (AI reasoning benchmark). Approached it like a startup product, significantly increasing data collection volume.

  • Migrated the system from JQuery to React

  • Improved grid manipulation UX, increasing object action usage (suitable for AI training) from almost non-existent to over 7 times per problem

  • Introduced gamification, events, and targeted ads to encourage users to solve problems voluntarily. Increased data collection speed from 2,000 in 6 months to 10,000 in 3 months

  • Published a paper as a co-first author at IJCAI 2024

  • Technologies: React, TypeScript, i18next, NestJS, Prisma

Introductory RL Study

2024. 01 – 2024. 02

Studied the basics of reinforcement learning through Coursera's Reinforcement Learning Specialization.

  • Studied the basics of reinforcement learning (– Q-Learning)

Startup Sigor Inc.

Web Developer

2022. 02 – 2023. 12

Doggly Frontend

2022. 08 – 2023. 03

Developed the frontend for Doggly (with over 2000 users), a shopping mall that measures dog sizes using AI and recommends fitting clothes.

  • Developed most of the frontend for the shopping mall, seller page, and admin page; worked on responsiveness and multilingual support; and handled maintenance

  • Developed the PWA & Jetpack Compose & SwiftUI WebView app and integrated the payment module

  • Introduced logging for error post-processing, enabling contact with users within 30 minutes in case of payment errors

  • Increased user conversion rate (product view) by over 40% through UI A/B testing over a month

  • Technologies: React, TypeScript, Styled-components, Jetpack Compose, SwiftUI, i18next, PWA

Rush Hour

2023. 05 – 2023. 12

Developed the frontend/app/ML for Rush Hour, a service that informs users about subway congestion and available seats.

  • Developed the initial version with React and WebView, then migrated to Flutter

  • Logged all user actions using GA, Amplitude, Smartlook, and a custom logging system

  • Developed AI to predict congestion levels at each station and time zone using subway congestion data

  • Technologies: React, TypeScript, Jetpack Compose, SwiftUI, Flutter, Firebase

Improving Productivity for Planners/Developers

2022. 02 – 2023. 12

As a developer at a startup, I focused on automating planners' tasks and improving developers' productivity. As a member of the early startup team, I handled various tasks including booth operations at expos and pet fairs, user interviews, and meetings with partners.

  • Improved CI/CD speed (from 3 minutes to 1 minute) by introducing zero install and Vite

  • Developed a web-based analytics tool focused only on the features needed for Rush Hour in response to feedback that Amplitude was difficult to use. Provided user flow, funnel, and retention analysis functionalities

  • Automated repetitive tasks of planners such as product approval and log extraction

  • Integrated mono-repo with Yarn Workspaces to enhance frontend reusability

  • Technologies: React, TypeScript, Amplitude, AWS Lightsail, AWS Lambda, Docker, FastAPI, Pandas, mono-repo

GIST Student Council Information Bureau

Senior Member; Team Lead (H2 2024)

2022. 09 – Now

Ziggle Frontend

2023. 04 – 2025. 02

Serving as team lead and developing the frontend for Ziggle, GIST's integrated notification platform.

  • Developed the initial version with React and migrated to Next.js

  • Introduced various technologies including GraphQL and Apollo Client

  • Maintained a DAU of 200 users, representing 25% of the undergraduate population

  • Technologies: React, TypeScript, Next.js, Tailwind CSS

Frontend Core

2022. 09 – Now

Focused on improving the frontend technical level within the team.

  • Introduced i18n for multilingual support

  • Adopted the Atomic pattern and documented with Storybook

  • Automated E2E testing using Playwright

IdP Frontend

2024. 03 – 2024. 12

Maintaining the frontend of the IdP, the OAuth2.0 authentication system used at GIST.

  • Frontend maintenance

  • Technologies: React, TypeScript, Tailwind CSS

Wing: GIST Computing Club

Senior Member

2023. 03 – 2024. 12

Wing Session

2023. 03 – 2024. 12

Conducting development knowledge exchange and sessions at Wing.

  • Led a session on introducing mono-repo

  • Led a session on React.cpp

  • Led a session on writing code like a UI library: Compound Components

GDSC-GIST

2nd Core, Tech Team

2022. 05 – 2023. 07

Study

2022. 11 – 2023. 05

Led the frontend study in the internal development study at GDSC. Created and conducted the overall curriculum. Also participated in the JS Pattern study, where I studied design patterns and applied them to a side project that created a C++ based React.

  • Frontend study lead

  • Participated in JS Pattern study

GIST Mini Solution Challenge

2023. 04 – 2023. 04

A hackathon organized by GDSC-GIST within GIST. A total of 60 people participated.

  • Hackathon preparation & accounting

GPERI

Research Member

2023. 07 – 2024. 03

C Language Education

2023. 07 – 2023. 08

Conducted C language education for the members.

  • Conducted C language education

  • Technologies: C

Rocket Communication Development

2023. 07 – 2023. 11

Attempted to develop rocket communication, but it was not successful.

  • Developed flight software

  • Developed communication system

  • Technologies: C, Zigbee

Side Projects

is-an.ai

2025. 05 – Now

A free domain service for AI services. Improving access for AI agents, with the goal of inclusion in the Public Suffix List (PSL).

  • Frontend Development

  • Research

Trax Music: GitHub for Musicians

2024. 09 – 2026. 03

A Git-like version control service for collaboration in music production. Supports feedback on specific sections of a track.

  • Frontend Development

  • Technologies: React, TypeScript, WaveSurfer.js

React-Atoms: React Component Library

2023. 07 – 2023. 07

This is a library of React components that I frequently use personally, published on npm. It is currently being used in this resume and my personal webpage.

  • Developed the library following the Atomic pattern, bundled with RollUp, and published on npm

  • Documented components using Storybook

Dutch: OOP Final Project

2023. 04 – 2023. 06

A Dutch Pay platform written entirely in C++ for both frontend and backend. It utilizes Emscripten to transpile C++ into Web Assembly. Presentation Slides

  • Implemented key React functionalities such as navigator, state, router, function components, and styles in C++

  • Developed a framework similar to React and Flutter

  • In browser environment performance tests, it showed approximately 110 times faster performance than React in hash repetitive calculation tasks

Speaking Melona: Convenience Store Food Talks

2024. 05 – 2024. 05

This service sets personas for convenience store foods and generates conversations between them when their barcodes are scanned. The challenge was barcode recognition in the frontend.

  • Frontend Development

  • Backend Development

  • Prompt Engineering

  • Technologies: React, TypeScript, FastAPI, PostgreSQL