MARK-E is a robot arm that draws your portrait. You stand in front of the camera, an AI model turns your photo into a simple cartoon, and the arm draws it on a northstar card while you watch. You walk away with the card.
It’s how we bring AI and robotics to people who’ve never touched either. Nobody needs an explanation: you see a photo become a drawing and a machine draw it, in a few minutes, and you keep the result.
How it works
The AI makes the cartoon and the robot draws it. A model draws you in the style of a Notion avatar, with clean black lines and big friendly eyes, and gives you three versions to pick from. Software turns the lines into pen strokes, and a DOBOT Magician arm draws them on the card. It runs on our own fine-tuned image model.
Most of the work went into the details you don’t see: getting a pen to press on a card the same way every time. We went through a floating sleeve, a flexure holder, a V-block and a collet before landing on the holder in the booth now, and replaced a magnetic card clamp with a simple hinged dock.
Where it’s going
MARK-E made its first portraits at the Cerebras × Codex meetup in Brussels in September, then drew its way through the Port of the Future Summit in Antwerp. Now it travels with Polaris Mini, our host robot, which greets people and sends their photo to MARK-E’s queue.
Next we’re moving it to a smaller model that runs on cheaper hardware, so it can go to more events. Next stop is Tectonic in Ghent.
