Founded and designed AIAgentBenchmark, an independent web project helping people and small teams understand whether AI agents genuinely improve real work.

I shaped the product strategy, visual identity, information architecture, editorial direction, and frontend experience. The platform translates complex ideas about AI agent evaluation into practical guides, benchmark resources, research notes, evaluation tools, and workflow-based examples.

The project focuses on questions that matter beyond impressive demos and leaderboard scores: Can an agent produce reliable work? Does it save time? Where does it fail? What should people measure before trusting it with more responsibility?

I designed and built the website using React, TypeScript, Vite, Tailwind CSS, React Router, and Cloudflare infrastructure. I also developed the content structure, SEO-friendly page architecture, resource taxonomy, navigation system, and workflow-specific request flow.

AIAgentBenchmark is intended for people beginning to use AI agents, professionals adopting AI at work, independent workers, small business owners, and small teams looking for clear and practical ways to evaluate agent performance.

Key areas of work:

- Product strategy and positioning

- Editorial and content-led product design

- Information architecture

- UX and visual design

- React and TypeScript frontend development

- SEO-friendly website architecture

- Benchmark and evaluation resource curation

- Workflow evaluation frameworks

- Brand and launch asset development