Work
My work focuses on the layers between computation and human sensemaking, where interpretation, cognition, and power quietly get encoded. My practice combines research, design, and concrete implementation to explore these questions in situ.
I come from HCI and interface design — building systems where tiny interaction choices decide what people notice, remember, and believe. Over the last decade the questions I explore have broadened. Less “better UI.” More: “epistemic infrastructure” — what technological architectures support humans in staying oriented when AI increasingly mediates cognition?
The projects and concepts here are organized around four interwoven themes:
Divergence Engines
Divergence Engines
Substrates that resist collapse
Most systems optimize for convergence — the answer, the summary, the first plausible output. This track explores the opposite: substrates that resist premature closure.
When retrieval and generation systems optimize only for relevance and fluency, they quietly collapse into sameness. The first plausible answer becomes the only answer. Exploration narrows. Alternatives disappear. I build technology that creates productive friction — systems that maintain possibility space, let structure compound from interaction, and actively resist the pull toward premature consensus.
| Pattern | Description |
|---|---|
| Auto-associative workspaces | Systems that observe usage patterns and form connections organically, allowing relationships to emerge from how content is used rather than requiring manual maintenance. |
| Recursive retrieval | Approaches that treat retrieval as navigation rather than search — multi-pass, reflective, across different layers of abstraction. |
| Divergence primitives | Architectural elements that preserve alternatives: branching, forking, maintaining multiple trajectories rather than collapsing to a single path. |
| Emergent structure | Systems where organization arises from interaction rather than getting imposed from above — bottom-up over prescribed. |
Cognitive Cartographies
Cognitive Cartographies
Knowledge as territory
Knowledge as territory, not inventory. These projects explore how spatial thinking and embodied cognition can make high-dimensional information landscapes navigable.
Our minds naturally process information spatially — we’ve evolved wayfinding abilities over millennia that most interfaces ignore. I develop approaches for transforming complex information into navigable landscapes: multi-scale navigation where you can move from overview to detail without losing orientation, terrain metaphors that make abstract relationships tangible, and visual differentiation that triggers natural pattern recognition. Maps, not folders. Territories, not hierarchies.
| Pattern | Description |
|---|---|
| Cognitive cartography | Systems that map information spatially, leveraging innate abilities to navigate and understand spatial relationships. |
| Strategic visual differentiation | Visual variations that trigger natural pattern recognition, helping users identify relationships without cognitive overload. |
| Semantic zooming | Dynamic content transformation that adjusts detail based on focus level, allowing fluid movement between overview and detail. |
| Embodied interface design | Physical metaphors in digital spaces — positioning, distance, and proximity as carriers of meaning. |
Ephemeral Interfaces
Ephemeral Interfaces
Design as meta-discipline
When computation is cheap and context is deep, interfaces don’t have to be fixed objects. They can be temporary cognitive surfaces — assembled for a task, shaped by where you are in the loop, dissolved when done.
This track explores what happens when design becomes a meta-discipline: when we design the conditions for interface emergence rather than the interfaces themselves. Design systems become grammar; interfaces become sentences generated for the situation. What persists is not the UI but the substrate — traces, relationships, provenance. The challenge isn’t building adaptive interfaces; it’s maintaining coherence and continuity as surfaces constantly reshape around intent.
| Pattern | Description |
|---|---|
| Interface agnosticism | Systems that maintain consistent capabilities while adapting presentation across contexts and devices. |
| Context-aware presentation | Surfaces that dynamically adjust information density based on current needs and available attention. |
| Transient composition | UI elements that appear when needed and dissolve when their purpose is fulfilled — ephemeral but not chaotic. |
| Substrate persistence | Architectures where the underlying structure (provenance, relationships, traces) persists even as the surface transforms. |
Cognitive Sovereignty
Cognitive Sovereignty
Agency in distributed cognition
What happens to agency when we delegate cognition to machines? Not simple binaries — open vs. closed, human vs. AI — but the scaffoldings that allow for synergies.
As AI systems increasingly mediate how we find, interpret, and remember information, questions of agency become critical. Who controls what shapes your attention? Can you trace how a conclusion was reached? Do you own your externalized memory, or does it fragment across walled gardens? This track explores systems where humans stay oriented and empowered inside interconnected environments — where collaboration (human-AI, human-human) compounds rather than fragments, and where transparency isn’t just visible but actionable.
| Pattern | Description |
|---|---|
| Workflow transparency | Systems where the reasoning process is visible and inspectable — you can trace how outputs were generated and what shaped them. |
| Provenance tracking | Architectures that preserve the chain from source to claim, making it possible to interrogate confidence and origin. |
| Context portability | User-owned infrastructure that travels across tools and providers, rather than fragmenting into platform-specific silos. |
| Selective revelation | Granular control over what’s shared and with whom — not opacity, but curated transparency. |
I’m looking for collaborators and teams working on epistemic infrastructure, human-AI interaction, and knowledge systems. If these problems resonate, let’s figure out how we can work together.