Built with Claude
We build Zaparo with Claude.
Claude, Anthropic's AI model, is part of how we work every day. This page says plainly where we use it today, where we do not yet, and where we want to take it.
Today
How we build with Claude today.
Three places, all of them running now.
-
Every line of code
Every app and game, every shared module, every server tool and this site itself is written and maintained with Claude Code. We set the direction and decide what ships; Claude Code writes the code, runs the builds and the tests, and fixes what they find.
-
The content our games ship
Our content factory has Claude write the word lists, quiz questions, characters and dialogue lines our games use, in English and in Turkish, each written in its own language rather than translated. A second Claude pass checks the work: it fact-checks quiz questions, drops words that do not fit a theme or are not right for children, and rejects lines that read like a translation. Only what passes is kept. Puzzle levels come from solvers we wrote with Claude Code, and no level ships until its solver proves it has exactly one solution.
-
Promo videos, written as code
Our promo videos are made with klip, a video tool of our own. Claude Opus writes each video as code, the picture and an original soundtrack alike, and klip renders it. The promo videos for our apps and our video ads are made this way.
Not yet
Our apps do not offer AI features yet.
We build our products with Claude, but the people who use them do not meet Claude inside them. None of our apps or games offers an AI-powered feature to its users today. Every word, question and line you see in them was written and checked before the app shipped, not generated while you use it.
Next
Next, we rework our apps with AI.
Soon we will go back through our apps and rework each one with AI, so it does more and asks less of the people who use it. We want to build that with Claude: the model that already writes our code and our content is the one we trust to work for our users directly. We will put AI where it makes an app genuinely easier to use, not where it only makes a feature list longer.
Contact
Questions about how we build?
Write to us about how we work, or about building something with us.