Federico Cohen Correa

Led design and delivery of 10+ technically complex digital products for US and LATAM startups — SaaS platforms, marketplaces, and AI-driven systems — owning end-to-end process from system architecture through production deployment.

• Led end-to-end design and delivery for 10+ products and multiple 0→1 MVPs — from problem definition and system architecture through production deployment.
• Designed role-based permission architectures and shadcn-based design systems for products with high operational complexity, ensuring direct parity between design and frontend implementation.
• Vibecoded production features directly in Next.js using AI-assisted workflows, shipping design decisions in code and eliminating traditional handoff cycles.
• Leveraged AI across research, product definition, prototyping, and iteration — increasing speed and decision quality throughout the entire product lifecycle.

Co-designed and vibecoded a 0→1 marketplace MVP end-to-end — built in 2 months on Next.js, deployed on Vercel, with Supabase (Postgres via Vercel integration), Clerk auth, and Google Maps.

• Designed full product architecture, service taxonomy, UI system, and core marketplace interaction model.
• Built 50% of all production features directly in code using Next.js, Cursor, and AI-assisted workflows.
• Fully integrated Vercel Blob (asset pipeline), Supabase via Vercel integration (Postgres), and Clerk (auth) — owning the complete infrastructure layer end-to-end.
• Defined early validation metrics with client marketing: listing creation rate, contact initiation, and return visit signals.

Led the full architectural redesign of a multi-role operational SaaS — from a no-code MVP to a scalable, role-based permission system rebuilt in Next.js with a shadcn design system.

• Designed a role-based permission architecture mapping explicit authority, visibility rules, and decision boundaries across multiple overlapping operational roles.
• Re-structured complex multi-stage workflows into modular, lifecycle-driven systems supporting regulatory processes and structured data traceability.
• Built with Next.js and a shadcn-based design system ensuring direct component parity between design and frontend implementation.
• Shipped production platform now supporting 81 championships and 43k+ users across 5 LATAM countries.

Explored and prototyped AI-driven legal intelligence workflows for a venture-backed predictive litigation platform used by general counsels and litigation finance teams.

• Designed and prototyped a GPT-powered deep research tool to identify high-probability litigation cases for funding evaluation.
• Rapidly prototyped alternative product directions, including executive-level dashboards and litigation portfolio tools with predictive case insights.
• Shaped predictive analytics workflows that connected AI outputs such as outcome probabilities and settlement projections to concrete legal decision points.
• Partnered with product and legal leadership to prioritize features in a high-velocity startup environment focused on rapid validation and iteration.

Led platform restructuring and interaction redesign for a creative review and approval platform used by retouchers, art directors, and production teams.

• Redesigned core interaction flows and UI components to improve clarity, usability, and feedback velocity in multi-round asset review workflows.
• Structured and scaled the design system with remodeled and newly defined components supporting evolving product capabilities.
• Defined and documented a robust role and permission architecture to support diverse profiles and secure access patterns.
• Established optimized handoff principles and delivery standards to increase design-to-code efficiency and reduce rework.

Other projects