We Turn Knowledge
Into Meaning.

We build and run web services that turn scattered data into results anyone can read.

About

WTKIM stands for "We Turn Knowledge Into Meaning." We build products that take data that already exists but is hard to read — survey responses, National Assembly records, YouTube subscriptions — and turn it into results anyone can understand.

WTKIM is a sole proprietorship registered in the Republic of Korea, founded in August 2026 by Wontae Kim. It is an AI-native studio that handles everything from planning and design to development and infrastructure. Anthropic's Claude is our core AI model, and our products are built on Next.js, NestJS and PostgreSQL and run on our own servers.

The company is currently run by its founder, and we plan to grow the team as our services grow.

Products

Products we currently operate or are building.

Survey platform

Beta
nas.polls.co.kr ↗

Upload a Word (DOCX) questionnaire and the platform imports its questions, scales and branching logic, then handles distribution, response collection, live monitoring and tabulation in one place.

  • Automatic DOCX questionnaire parsing (questions and branching)
  • Login-free token links for collecting responses
  • Live dashboard with cross-tabs and drop-off rates
  • Claude converts uploaded documents into question structures (planned)
  • Claude analyzes responses and drafts reports (planned)

Legislative record app

In development
vote.polls.co.kr

Shows bills introduced in the Korean National Assembly as cards. Users record their own position on each bill and compare it side by side with how each party actually voted, based on official Assembly records rather than news commentary.

  • Bill and vote data from the National Assembly Open API
  • Agreement score between your positions and party votes
  • Summarizes the issues without making the call for you
  • Installable PWA, no app store needed

chchu

Live
chchu.com ↗

Builds a viewing-taste profile from your YouTube subscriptions and lets you compare overlapping channels and recommendations with friends through a share link.

  • Google sign-in (requests read-only access to subscriptions)
  • Claude analyzes subscriptions to generate a taste report and interest tags
  • Compare channels and recommendations with friends

How we work

Ship small, ship early

We build the essential features first, put them in front of real users, and refine based on how they respond.

AI at the core

We use Claude to shorten work that used to take people a long time, such as document parsing, response analysis and report writing.

Data with a clear source

We use only public APIs and data users explicitly allow, and request only the permissions we need.

Contact

For product questions, partnerships, development projects or joining the team, email us.