Tiana Liu

AI product · Brand systems · 2026

Giving an AI filmmaking product a human point of view.

I turned a generic AI-generated identity into a deliberate multi-platform brand system-and defined where AI strengthened the work versus where human judgment mattered more.

RoleBrand Designer
TeamFounder + stakeholder
ScopeStrategy to delivery
ToolsFigma + AI workflows

The product promised cinematic originality. Its identity looked machine-made.

Loovie turns written prompts into films using multiple AI engines. Early feedback described its original AI-generated logo as generic and difficult to trust. That perception directly contradicted the product promise.

Design question: how can an AI-native brand feel creative and specific-not statistically average?

I used AI as a reviewer, not the creative director.

Initial generation tests produced polished but forgettable options. I moved the creative core back to research and human-led concept development, then used AI selectively for technical critique.

Human-led direction

Studied film, creator tools and mobile brands to define a cinematic, approachable tone.

AI-assisted refinement

Used structured critique for spacing, proportions, line weight and small-size legibility.

Stakeholder alignment

Framed the decision as a product trust problem rather than a subjective style preference.

Loovie brand identity exploration

One identity, designed to survive real product constraints.

I translated the approved direction into a usable system covering app icons, in-product use, website assets, social formats, colour, typography, clear space and minimum-size rules.

  • Platform-specific specifications for iOS, Android and web
  • Developer-ready usage rules across light and dark surfaces
  • AI-assisted documentation built from decisions I had already defined

The strongest AI product judgment is often knowing when not to automate the decision.

What this demonstrates: AI workflow design · product judgment · systems thinking · stakeholder communication