Selected work 02
Role
UX/UI Designer
Category
AI / SaaS / Creative tools
year
2025
Type
Product design / MVP
Context
Visuella is an AI-powered visual platform for marketers, creators, and small brand teams who need campaign-ready visuals without learning prompts, editing broken layouts, or rebuilding the same content for every format.
Brand direction
The product started from a simple tension: AI can generate almost anything, but busy non-technical users still struggle to get visuals that feel consistent, branded, and usable. The goal wasn’t to make “another AI generator”. It was to make AI feel like a calm visual assistant.
Problem
Creating visuals is easy. Creating branded visuals isn't.
01
Users don’t always know how to describe what they want visually.
02
AI visuals often look good once, but don’t stay aligned across campaigns.
03
One idea still needs to become a post, story, banner, and ad format.
04
Small changes can break the layout, product angle, or visual balance.
TARGET USERS
The core users are brand marketers, content creators, small business owners, and social media leads. They are visual enough to care about aesthetics, but often don’t have time, budget, or technical confidence to work with advanced AI tools.
To understand how creative teams actually use AI visuals, I conducted 7 interviews with marketers, brand designers, and content creators.
*Remote interviews with marketers, brand designers, and content creators
RESEARCH
Built specifically for fashion, beauty, and DTC brands. Offers visually strong product renders and branded scenes, but still expects users to think like designers. Creating the right result often requires experimentation, moodboards, and manual adjustments.
Fast and approachable for simple product imagery. Users can quickly place products into ready-made scenes, but personalization remains limited. Outputs often feel template-driven and less aligned with unique brand identities.
Combines generation, editing, and publishing in a familiar ecosystem. Accessible for a broad audience, yet the experience is heavily editor-focused and generated content can feel generic without significant customization.
Produce high-quality visuals and videos with impressive creative freedom. However, both rely on prompting expertise and offer limited control over brand consistency, layouts, and repeatable marketing workflows.
The gap wasn't image generation. The gap was brand consistency. Existing tools help users create visuals. Few help them create visuals that stay aligned with brand guidelines, adapt across formats, and remain easy to edit without breaking the final composition.
USABILITY TESTING
Before moving into high-fidelity design, I tested the core experience with 8 participants using Maze. The goal was to understand whether users could successfully navigate the workflow, understand the product logic, and complete key tasks without guidance.
The tests were conducted on wireframes, allowing me to validate the experience early and identify usability issues before investing time into visual design.
01
Participants often hesitated when asked to create visuals from scratch. Many expected suggestions, templates, or guidance before taking action.
02
While automatic generation was appreciated, participants wanted confidence that important elements wouldn't change unexpectedly during edits.
03
Most users preferred predictable, brand-aligned results over endless creative variations.
DESIGN RESPONSE
The research revealed a common pattern: users didn't need more generation options. They needed more confidence, guidance, and consistency throughout the workflow. To address these needs, I focused on three core product directions.
Generate campaign-ready assets for different channels without rebuilding visuals from scratch.
FINAL SOLUTION
The final product combines AI-powered image generation with a structured workflow focused on brand consistency, visual quality, and ease of use. Rather than relying on prompt engineering, users are guided through a visual-first experience built around moods, references, and brand assets.
MARKETING EXPERIENCE
Alongside the product experience, I designed a marketing website focused on explaining the product through outcomes rather than technical AI terminology. The goal was to immediately demonstrate what users can create, how the workflow works, and why it feels different from existing AI tools.
Reflection
Working on Visuella reinforced an important lesson: the challenge is no longer generating visuals. The challenge is helping users trust, control, and reuse AI-generated content within real workflows.
The strongest opportunity wasn't creating more images. It was helping teams create visuals that remain consistent across campaigns, formats, and iterations.
Available
© 2026 Daria Perepadia. All Rights Reserved.