Raul Testa — Lead Product Designer & AI-Native UX Architect

The challenge

Multi-chain traders don't share one workflow. The existing terminal shipped a fixed dashboard, so every user got a layout built for someone else — and the navigation ate screen space that should have belonged to live market data.

What I did

1. Ran user interviews and persona work alongside the Product Owner to map how traders actually monitor positions across chains.
2. Designed an "Add Window" system from scratch — a 4-step flow where any user composes their own multi-chain view, with favorites, token switching and in-place detail expansion.
3. Slimmed the navigation architecture to give live data back the real estate it needed.
4. Replaced the old sidebar with a slim left nav, freeing a large block of pixels for the widget dashboard.
5. Rebuilt mobile around the same window logic instead of shrinking the desktop grid — bottom nav plus a central CTA.
6. Ran colour-blind alternative studies (protanopia, deuteranopia, tritanopia) on active/selected button states before locking the palette.

The outcome

One coherent window model across desktop and mobile: users configure the workspace once and recognise it everywhere. Every decision was validated against AAA accessibility standards before handoff to engineering in Figma.

The challenge

Six products, one ecosystem, six different audiences — from institutions moving real-world assets to people opening their first wallet. Left alone, each would have drifted into its own visual language and its own navigation rules.

What I did

1. Standardized on responsive core containers so desktop and mobile always read as the same product.
2. Adapted navigation per audience: a data-dense top toolbar for Tashi Finance, a side nav for Kinetic to protect horizontal real estate, progressive disclosure for newcomers on Hover Market.
3. Built out Trad.Fi across 100+ screens — the largest Figma file I've produced — covering credit pools, asset onboarding and institutional reporting.
4. On Tashi Finance, turned the main Supply/Borrow navigation into a large centre-screen toggle so the most important figures were never more than one tap away, and moved secondary actions into a hamburger menu.
5. On Hover Market, concentrated navigation in a top toolbar and made the hamburger behave as a right-side slider — buying back vertical real estate for market data.
6. Learned the token economics behind BITS before designing wrapping, unwrapping and staking, so the flow matched what was actually happening on-chain.

The outcome

A suite that reads as one family without forcing one interface onto very different users. Shared containers cut redesign work on every new product, and the per-audience navigation kept first-time crypto users from bouncing on their first screen.

The challenge

Two audiences with opposite needs sharing one data model: citizens who want a personal wellness habit, and mayors who need population-level evidence before committing public money. Serving either badly would have sunk the other.

What I did

1. Co-created the concept as business partner, then ran the quantitative and qualitative research behind it.
2. Sketched both surfaces by hand before moving into Figma with the dev team, so the information architecture was settled before pixels.
3. Designed the public app around gamified health plans tied to a local-currency reward spendable with local merchants.
4. Built the official dashboard to surface BMI, activity, smoking and chronic-illness trends — plus a mobile version so policymakers can check it anywhere.

The outcome

One product, two clearly separated surfaces: citizens never see population data, officials never wade through consumer features. The Leiria pilot dashboard reported on 68,564 users and €350,867 invested across January–August 2022, alongside population indicators like 85% above recommended BMI and 82% tobacco use — the kind of figure a mayor can act on.

The challenge

The people SPEAK exists for are often new to a country, a language and the platform at once. Every extra form field, every unexplained term is a reason to leave — and volunteers, not staff, keep the classes running.

What I did

1. Drew the original storyboard by hand in 2012 — migrant arrives, finds a class, ends up with a multicultural circle of friends — and designed the product from it.
2. Created the brand, the web platform and the mobile app as one system.
3. Held the core task — enrolled and into a class — to four clicks or fewer.
4. Replaced self-declared skill levels with a short onboarding quiz that places each learner before they ever see a class list.

The outcome

SPEAK grew from a single NGO project into the largest private migrant-integration platform in Europe, now running across 12 countries including Nigeria and Bangladesh — on a flow that still takes under a minute to complete.