Based
Available globally · remote-friendly
I'm an IC product designer who works on the layer where users meet large systems — feeds, AI co-pilots, and the trust mechanisms in between. My job is to defend the visual restraint and product clarity that lets ambitious systems feel inevitable instead of imposed.
The case study I built for the xAI Product Designer role — proposing Grok as a layer over the timeline, not a destination beside it.
A product design proposal arguing that Grok shouldn't be a tab — it should be a layer over the timeline. Six surgical insertions, ordered by impact. One monochrome system. Zero new tabs. Spec doc, visual board, and a working interactive prototype.
Earlier projects across consumer AI, social products, and system design. Each named with the bet it argued, not the company that shipped it.
A teardown and redesign of consumer-AI onboarding flows. Argued that the first 90 seconds — not the model — determine retention. Shipped a new pattern: skip the chrome, hand the user a working capability inside their first sentence.
Reimagined bookmarking as a queryable personal archive, not a graveyard. Built the patterns for surface-level retrieval (search), mid-level synthesis (tags, clusters), and deep recall (conversational query) — and the trust scaffolding that lets each layer fail gracefully.
A design-system audit and rewrite for a product that had drifted toward decorative AI flourishes (gradients, particle effects, "shimmer"). Killed every non-functional treatment. Wrote the principles that survive scope creep — and the language the team uses to defend them.
A study and prototype on how AI products communicate confidence to users. Compared hidden, implicit, and explicit uncertainty patterns. Concluded — and shipped patterns based on the finding — that explicit confidence is the single largest moveable input on long-run user trust.