A visual identity that doesn’t try to look like AI.
As the first design and brand hire at Symbolica, I shaped an identity that refuses to take part in AI hype design — drawn instead from the mathematics itself. Category theory’s structure, repetition and controlled variation carry the rigor; bespoke illustrations from commutative diagrams bring the research to life; and motion is a living, code-driven system that evolves alongside the work.
View Official Brand GuidelinesBeyond pattern matching.
Most models predict the next token. Symbolica builds models that reason over structure — applying category theory to give machines a grammar for thought.
Intelligence, redefined.
Paper, ink & a single periwinkle.
Using formal computational logic, rather than gradient descent, Symbolica's models reason beyond current limitations.
A data point, in motion.
Arrows, self-loops and nodes trace a data point through the learning process — periwinkle as linework, never as a background.
A living motion system.
Real-time generative animations, driven by parameters rather than shipped as fixed assets — a code-driven library that evolves alongside the research.
Play with the asset. Drag the parameters — spawn rate, curvature, frequency — and watch the system render a new variant in real time.
Play with the asset
