Agentvet — Directory & Benchmark Platform

Project overview Product: AgentVet.ai Engagement: Product design and frontend build My role: Product design, UX/UI, frontend development Tools: Figma, HTML, CSS

The challenge The AI-agent market is crowded with marketing claims, inconsistent categories and little comparable evidence. Users need a way to discover agents while understanding what each one does, how it performs and why a rating should be trusted.

The approach I structured the experience around progressive confidence: browse a category, compare candidates, inspect rankings, then open a benchmark certificate. Dense data was treated as a product hierarchy problem rather than reduced to decorative cards.

The solution The interface combines search, category filters, agent profiles, rankings and scored benchmark certificates in one dark visual system. Large typography creates a clear entry point, while restrained color and structured tables keep comparisons readable.

The outcome AgentVet was delivered as a working responsive product surface at agentvet.ai. The design gives the platform a credible, evidence-led position and establishes reusable patterns for new agents, benchmark dimensions and categories.

Deliverables Information architecture, landing page, directory and search patterns, rankings, agent profiles, benchmark certificates, responsive UI and frontend implementation.