Case study
Revolico
Frontend engineering and product UI for search, listings, publishing flows, admin interfaces, and performance.
Revolico is a high-traffic classifieds marketplace for Cuba. I worked on React/Next.js UI for search, listings, publishing flows, contact paths, admin tools, feedback, and performance. The work supported faster pages, clearer flows, and more actionable user feedback.
- Role
- Frontend Developer + Product UI
- Period
- 2017-Present
- Ownership
- Frontend engineering and product UI across marketplace search, listing detail, publishing flows, contact paths, admin/product operation interfaces, performance improvements, feedback mechanisms, and testing.
25%
TTFB reduction
30%
stronger engagement metrics
1,000+
actionable user insights from feedback mechanisms






Problem
Keep a large marketplace searchable, fast, and understandable
A classifieds marketplace has to support many user intentions at once: browsing, searching, filtering, publishing, contacting sellers, managing listings, and navigating many categories.
In Revolico's context, the product also needed to remain fast and usable for users in Cuba, where connection quality can be a major constraint.

Users / context
Buyers, sellers, businesses, admins, and mobile users
Revolico serves buyers searching for products, services, housing, jobs, cars, technology, and general goods; sellers publishing ads; businesses seeking visibility; admins managing platform operations; and mobile users on web or app workflows.
Public descriptions emphasize direct buyer/seller contact, free ad posting, visibility, speed, broad marketplace categories, and optimization for Cuban connections. The interface work had to support both high-repetition marketplace browsing and focused mobile tasks such as filtering, publishing, viewing details, and contacting sellers.

Constraints
High traffic, broad categories, publishing speed, and modernization
Public marketplace pages needed to be reliable and performant across many categories, including vehicles, real estate, technology, jobs, services, home, family, and general goods.
The product had to support fast publishing, predictable search and filtering, admin moderation workflows, connection-quality constraints, and long-running modernization without breaking existing behavior.
Design decisions
Favor scanability, direct navigation, and performance-aware UI
Marketplace UI depends on repetition and confidence. Category navigation, listing title, price, context, images, and contact affordances needed to be easy to parse without decorative weight.
Publishing and admin flows needed clear form states, errors, progress, filters, permissions, and edit flows so the team could move quickly without creating operational ambiguity.
- Keep category navigation direct.
- Optimize listing scanability.
- Support quick publishing with clear form states.
- Treat admin UI as product UI.
- Keep interaction patterns lightweight for performance.

Interaction / UI system
Marketplace components for search, listings, publishing, and admin workflows
The repeated UI patterns show up in homepage and category navigation, search results, listing grids, ad detail pages, publish-ad flows, filters, sorting, pagination, admin user and listing management, moderation flows, promoted-ad surfaces, and mobile store screenshots.
The component work covered marketplace cards, buttons, inputs, dropdowns, breadcrumbs, pagination, image selectors, category taxonomy UI, admin table and filter patterns, and performance-aware image handling.

Implementation collaboration
React/Next.js frontend work for a high-traffic marketplace
I worked as frontend developer and product UI contributor on public-facing features and admin interfaces.
The implementation included React/Next.js, Apollo Client, GraphQL, Styled Components, SSR/SSG, SuperTokens, i18next, Jest, Cypress, Cucumber, feedback mechanisms, performance work, backend/frontend collaboration, and testing coverage.
Outcome
Performance, feedback loops, and product reliability
Confirmed outcomes include 30% stronger engagement metrics from iterative releases and feedback-informed product work, approximately 25% TTFB reduction through performance and backend/frontend collaboration, and 1,000+ actionable feedback insights.
The work also improved reliability through test coverage and supported React/Next.js development for a high-traffic classifieds marketplace.
Stack
