Encode: AI for Science is a fellowship founded by Pillar VC, connecting leading AI researchers with top UK academic labs to accelerate scientific discovery. We worked with Pillar to build a brand that captures both the precision of advanced computation and the creativity that drives research, resulting in an identity that feels technical, intelligent, and alive.

‍

The concept is rooted in embeddings and dimensional reduction — a process in AI that simplifies complexity into its most essential form. At the heart of this idea lies the dot, the fundamental unit of the Encode system. Just as embeddings transform high-dimensional data into meaningful representations, the brand builds from this single element to create a vast, adaptable visual language.

‍

From this foundation emerges a network of patterns, motion principles, and data-inspired compositions that visualise connection, collaboration, and transformation, the fellowship’s core purpose. The dot acts as the building block of the entire system, scaling from micro to macro applications. It shapes photographic treatments, defines motion behaviour, and drives the formation of intricate visual structures that evoke the flow of data and ideas. Through repetition and interaction, it becomes a flexible device capable of expressing both structure and fluidity, mirroring the nature of AI-driven discovery itself.

‍

The Encode logo unites these principles through dimensionality. Built from the shape of the embed symbol in programming, it expands across three axes and incorporates a subtle star reference, symbolising orientation, exploration, and scientific pursuit. The dot moves from the icon into the wordmark, becoming a linking element that binds the system together. The result is a mark that feels engineered yet fluid, reflecting Encode’s mission to translate the complexity of science into clarity and connection.

‍

The website was designed to introduce Encode’s first cohort and encourage applications to the programme. Its structure reflects the clarity and dimensionality of the brand system — organised, layered, and intuitive. Multi-level interactions bring depth to the experience, subtly referencing the concept of embedding and the visual language of AI. Motion and visual treatments extend directly from the brand’s core principles, using the dot and its dynamic behaviour to guide navigation, highlight content, and create a sense of flow. The result is a digital presence that feels both precise and exploratory, showcasing Encode as a hub of innovation and collaboration within the AI for Science community.