TickerPulse

What I did

• Aligned closely with the PM to define scope, priorities, and MVP direction.
• Translated product intent into clear UI structure and interaction flows.
• Rapidly prototyped using Figma Make to support fast feedback and iteration.
• Prepared clear design artifacts to support discussions and handoff with engineers.

Background

Exploring vibe coding with AI-generated UI design

This project explores vibe coding using Figma Make powered by Claude 4.5 Sonnet, marking the first time AI-generated UI design achieves human-level quality. Built purely from prompts and reference images without any manual design work, it reduces prototyping time from days to hours. By enabling teams to quickly prototype ideas and gather feedback through usability testing, vibe coding accelerates validation and iteration while allowing designers to focus on creativity and strategy instead of production.

Problem

Cluttered dashboards create cognitive overload

Financial traders face cluttered, fragmented dashboards that force constant switching between analysis, news, and execution tools. Poor visual hierarchy and lack of AI guidance create cognitive overload, slow reactions, and missed opportunities during fast market shifts.

Solution

Unifying market insight and action in one interface

TickerPulse unifies market insight and action in one "Alert-Context-Action" interface. With clear visual hierarchy, progressive information flow, and AI-powered insights, it reduces platform switching and decision time, helping traders move from signal to execution in under three clicks with clarity and confidence.

Reflection

What I learned

I built the initial layout in Figma, then used Claude 4.5 Sonnet in Figma Make to refine details — iterating through prompts grounded in project context and design judgment to reach the final design. The prompt quickly reached its limit, and there are still many features worth exploring. For the MVP, the core functions worked well after testing with stakeholders. Moving forward, the prototype flow will be co-created with AI, making it essential to describe intent precisely. In this process, the designer acts as a taste gatekeeper who guides the AI toward the right vibe and experience. While developer support is still needed for API integration, the line between designer and developer is becoming increasingly blurred, showing how anyone can rapidly prototype and validate ideas with AI.