Lumen AI

Project Overview

Lumen AI is a marketplace where businesses discover, compare, and deploy specialized AI agents (e.g., customer support bots, data analysts, content generators). The challenge: users felt overwhelmed by technical specs and struggled to identify which agents fit their actual workflow—leading to slow decisions or abandoned searches.

The task was to design a clean, data-driven interface that made it easy for users to compare AI agents based on their specific needs and use cases. I focused on creating intuitive navigation, clear categorization, and visual cues to help users quickly understand the capabilities of each agent and how they could fit into their workflow.

Design Process

1. Research: Reviewed competitor marketplaces; identified pain points—users skipped agents with unclear value propositions.
2. Structure: Mapped core user flows—browse → compare → deploy—and simplified steps (e.g., removed unnecessary "request demo" gates for self-serve agents).
3. Wireframes: ested layout options for data density; settled on card-based browsing + expandable comparison table.
4. UI Design: Built in Figma using a minimal component library (buttons, data cards, filter chips). Applied WCAG contrast ratios for readability.
5. Validation: Checked key screens on real mobile/desktop devices to ensure touch targets and text sizing worked in practice—not just in Figma.

Result

• Users can identify a relevant agent in less than15 seconds (vs. scanning long spec sheets)
• Components are production-ready: named layers, Auto Layout applied, 8px grid enforced
• Design scales cleanly as new agents are added—no layout rework needed