AigentGallery

Context

AI agents are becoming increasingly specialised, but discovering the right agent is still difficult. New tools are scattered across product launches, social platforms, directories and individual websites, making it difficult to understand what an agent does, which problem it solves, and how it compares with other tools. I built AIgentGallery as a curated AI agent directory — a focused discovery experience where people can explore newly launched agents, browse them by category, and quickly understand what each product is designed to do. The goal was to make AI agent discovery feel simple, visual and useful rather than overwhelming.

Contribution

I took AIgentGallery from product concept to a working product, combining product design, UX/UI design and frontend development.

• Product strategy and UX
• Information architecture
• UI design
• Responsive web design
• Frontend development
• Agent submission flow
• Moderation experience
• Category and discovery system
• Production build

Constraints

• Constraint 01

  Make a growing AI category easy to navigate

  AI agents can serve completely different purposes, from coding and research to sales, productivity and automation. The interface needed a simple categorisation system that allowed users to quickly understand where an agent belonged without introducing unnecessary complexity.

• Constraint 02

  Keep the directory curated

  An open submission system could quickly lead to low-quality, duplicate or irrelevant listings. Rather than allowing every submission to appear immediately, the product needed a structured submission and moderation flow so the directory could remain focused and useful.

Key decisions

• Decision 01

  A launch-board model instead of a traditional directory

  I wanted AIgentGallery to feel like a place to discover what's new rather than a static database of tools. The homepage therefore puts recently launched and ranked agents at the centre of the experience, giving users a reason to return and discover something new.

• Decision 02

  Category-first discovery

  Instead of relying entirely on search, I introduced clear categories based on the job an AI agent performs. This creates a faster path from a broad need to a relevant set of products.

• Decision 03

  Curated submissions instead of unrestricted publishing

  The submission flow was designed around moderation rather than instant publishing. This gives AIgentGallery control over the quality of the directory while still allowing new products to be submitted through a structured process.

Outcomes

• A complete AI agent discovery platform

  AIgentGallery launched as a working product rather than a static directory concept, with agent listings, categories, ranking, individual agent pages and a submission experience.

• Designed and built as a solo product

  I took the project from product concept and UX through interface design, frontend development, submission, moderation and production launch. The result is a focused AI agent directory designed to make discovering emerging AI products faster and easier.