Neurometric
Turning AI benchmarking into a vibrant playground.
Neurometric is an AI startup with a simple but provocative idea: put large language models in the ring, give them the same task, and see what happens. It’s part analysis, part experiment, part spectacle—comparing how different AIs think, write, solve, and respond. My role was to shape the interactive experience, visual identity, and brand system for this concept, turning what could feel like a technical benchmarking tool into something people actually want to explore. The challenge was balancing intelligence with energy—making it credible, but never cold.
The visual language pulls from spirographs, neural pathways, and looping systems—forms that feel alive, slightly unpredictable, and always in motion. Bright, expressive color cuts through a clean, modern foundation, while a cast of playful icons and pictograms act like a shared language between human and machine. Everything is simplified just enough—language, layout, structure—so the complexity underneath can breathe. It doesn’t explain the models so much as let you watch them think.
Closer to a live experiment than a spreadsheet.
The experience itself leans into that sense of discovery. Instead of static comparisons, outputs unfold side-by-side, revealing quirks, strengths, and unexpected differences in real time. Subtle motion, responsive feedback, and layered interactions make each comparison feel less like reading a report and more like witnessing a matchup. It’s analytical, but with a pulse.
To keep it all cohesive, I built a flexible set of brand guidelines that define everything from color logic to motion behavior to tone of voice. The system is designed to scale as the product evolves, without losing its personality. Neurometric ultimately lives in that sweet spot between precision and play—where serious technology meets curiosity, and where comparing AI doesn’t just inform you, it pulls you in.