LogicHum is a shipping studio where creative writing and machine learning work side by side—making games, apps, and tools people can actually use.

LogicHum started with a simple irritation: most interactive products either sound good or work well—rarely both.

We came in from two directions. One of us grew up in translation—language, place, expectations—and learned early that meaning changes when you move it. That immigrant muscle became a craft: hold two truths at once, and don’t lie to make them match. The other obsession was systems: rules, feedback loops, the way a small constraint can produce a world.

At first, it looked like separate work: writing on one side, models and prototypes on the other. Then the overlap got loud. We kept noticing the same gap: narrative wants nuance; software wants structure. We didn’t want to “AI-wash” voice into paste, and we didn’t want story to be a thin skin over mechanics.

So LogicHum became a practice: write the world clearly, build the system honestly, then iterate until the experience feels authored and responsive. The machine isn’t the author. It’s the partner that helps us test, vary, compress, expand—without losing the spine of the piece.