My role: identity, photography, product design, development and catalogue setup.
The Problem
A small saree label needed a coherent storefront and a maintainable system behind it. The source audit found undocumented tokens, repeated motion values and readability and focus gaps.
The limits
One designer and developer, the existing brand’s visual identity, and a live storefront. Library-dark is a demonstration; checkout remains an unreleased preview.
What I built
Primitive → semantic → component tokens. One versioned JSON source, an alias-resolving compiler, CSS outputs and a Figma library with parity-checked scalar values.
What happened
+395% orders a month, from 3–4 to about 17 through the website, by the founder’s count. Built, not shipped: 206/206 tokens documented and 93 → 0 hand-typed timings. Growth is not attributed to the design system alone.
My role: product design, the prototype, and the internal build’s interface and backend, with the repository team, IT and senior producers.
The Problem
Teams pick images before anyone has read the licences, so problems turn up mid-production and work gets redone.
The limits
Clario only reads, and takes limited data. Missing evidence must stay uncertain, and a person keeps every commercial decision. It runs on rules, with no AI.
What I built
One check before the ads get built. It reads every image’s dates and channels and marks it cleared, review or at risk. The public prototype runs on made-up records.
What happened
More than 15 campaigns in under two months, with dozens of producers. It found problems before production, on images reused from older campaigns. Rework savings aren’t measured yet.
