Fig. 01 — Mdici Prototype, scouting flow.
Fig. 02 — Player Profile view, scout dashboard.
Mdici is a sports-tech product built around one question: how do football academies, scouts, and clubs find the next generation of players before the rest of the market does?
The product runs on two surfaces. A social-style mobile app, where young players publish performance content, training clips, and match highlights — the place where talent makes itself visible. And a scouting dashboard, where football schools and professional scouts review that content, track player development, and produce structured reports on the athletes they’re following.
When I joined, the social app already had an early audience but the scouting product didn’t exist. The brief was to lead it from a blank page to a working platform that real academies would adopt.
Mdici is the project where my three disciplines — product design, team leadership, and front-end engineering — collapsed into a single role.
On the social app, I led a team of four designers, owning hiring, weekly direction, design reviews, and the standards the team shipped against. I set the visual and interaction system that the mobile product evolves inside, and ran the rhythm that kept four people aligned across feed design, content moderation tools, profile structure, and the social mechanics that drive young athletes to publish.
On the scouting dashboard, I owned the product from research to production code. I designed the full interface in Figma, then implemented it myself in React — working day-to-day with the backend team on data contracts, with the machine-learning team on how player metrics and highlight detection surface inside the UI, and with marketing on how scouts and academies discover and onboard onto the tool.
The dashboard couldn’t be designed in a vacuum. Before any wireframes, I ran a focused research engagement with a partner football academy in Tarragona that agreed to act as our reference user.
Across multiple sessions on the ground, I worked with coaches, scouts, and technical directors to map exactly how they currently identify, evaluate, and track players — what gets written down, what gets remembered, what gets lost, and what they wish they could see across multiple training sessions and matches.
Three priorities came back consistently. They wanted to see player highlights without scrubbing through full match videos. They wanted to compare players across the same evaluation criteria, not subjective notes. And they needed a structured report at the end of each cycle — something they could share internally or with parents and clubs without rewriting it from scratch every time.
Those three priorities became the spine of the dashboard.
I designed the scouting dashboard around the three priorities the Tarragona academy gave us: highlights, comparison, reporting.
The highlights layer is powered by the platform’s machine-learning pipeline, which automatically tags the strongest moments inside any uploaded clip. I designed the UI patterns that surface those tagged moments — scrubbing, filtering by event type, jumping directly to the play that matters — so scouts spend their time evaluating, not searching.
The comparison layer turns subjective evaluation into a structured grid. Each player carries a consistent set of metrics, and scouts can stack two or more athletes side by side under the same criteria. This was the feature the academy adopted fastest: it replaced a workflow that used to live in spreadsheets.
The reporting layer compiles everything into a structured player report — ready to be shared with coaches, clubs, or families. It’s the artifact the scouting workflow had always been missing.
I designed the system in Figma, built the design tokens and the component library that the engineering team uses, and implemented the production front-end in React.
Mdici only ships because design, backend, machine learning, and marketing operate against the same plan.
I worked closely with the backend team to define the data contracts that feed the dashboard — what a “player” object actually contains, how a clip is structured, where moderation lives, what the reporting layer reads from. With the machine-learning team, I designed how the model’s outputs — auto-tagged events, suggested highlights, performance scores — show up in the UI in ways that scouts can trust and override. And with marketing, I made sure the onboarding surfaces inside the product matched the story we were telling academies on the outside.
This is the level of cross-functional coordination the role demanded daily — and the one I’ll keep doing on the next product.
“The work was equal parts designing a product, leading a team, and writing the code that proves the design holds up.”
Mdici was the project where I stopped thinking of design, team leadership, and front-end as separate roles.
The dashboard works because the research is honest, the team is aligned, and the production code matches the design intent — and those three things are inseparable. When the design team and the engineering team work against the same definitions, decisions move five times faster. When the research is grounded in a real academy instead of an internal hypothesis, the product feels obvious to the people it’s built for.
The harder lesson was about managing a creative team while staying close to the work. I learned to defend my designers’ calendars, to push back on scope, and to invest the same care in weekly one-to-ones as in the product itself. Most of what shipped well, shipped because the team had the conditions to do their best work.