UX / Product Design
A WhatsApp-native AI marketing engine that turns product ideas into high-converting video ads delivered in under 60 seconds.
Project Overview
My Role &
Responsibilities
UX/UI & Product Designer (Solo designer, lean startup) · Experience Architecture
Conversation Design · AI System UX
Landing Page · Brand
001
Problem Statement
Small and mid-sized businesses struggle to create high-quality ads quickly.
Studio shoots are expensive
Agencies take too long
Multiple tools create friction
Non-technical founders feel overwhelmed
Slow creative turnaround delays campaigns
The core issue wasn’t just ad creation, it was complexity.
002
Soluton
Built Zebvo AI - a WhatsApp-native AI marketing engine that:
Generates multiple video and graphic ads in minutes
Delivers creatives directly on WhatsApp
Reduces dependency on multiple tools
Simplifies AI interaction for business owners
One system. Faster execution.
Goals
Reduce creative turnaround time and simplify the entire ad creation journey while maintaining professional output quality.
(02)
(Design Decisions & Solution)
© 2026
Experience Architecture
Designed a frictionless 7-stage flow - from ad discovery to WhatsApp delivery - entirely without a dashboard or login. Every touchpoint was mapped to reduce psychological friction and keep the user moving forward, not second-guessing the system.
Conversation & Data Structuring
Instead of open-ended prompts, I designed guided questions that extract AI-ready structured inputs - brand details, USP, tone, audience - without overwhelming non-technical founders. The form isn't data collection. It's intelligent input shaping.
AI System Orchestration
Designed how structured inputs flow through a multi-layer AI pipeline - interpreting product signals, generating scene-based script variations, and filtering only the top outputs for delivery. The goal was controlled creative exploration, not AI randomness.
Userflow
This diagram maps the full end-to-end system that powers Zebvo AI — from the moment a user submits their product details to the final creative landing on their WhatsApp. Every decision point, processing layer, and output path is designed to run automatically, with no manual intervention at any stage.
Experience Validation
Brands completed the WhatsApp onboarding without friction or manual guidance. No major drop-offs in the structured form submission. End-to-end automation ran reliably within ~60 seconds no manual coordination required at any step.
Business Validation
Achieved a 2× improvement in ROAS after messaging simplification. Acquired an early cohort of paying users with sub-target CAC demonstrating viable unit economics. Shifted from early-stage loss to profitable acquisition within the first campaign cycle.
Outcome
The WhatsApp-native flow ran end-to-end with zero drop-offs in onboarding. Automation delivered creatives reliably within ~60 seconds across every user - no manual coordination needed at any step.
Campaign performance shifted significantly once messaging was simplified - achieving 2× ROAS and moving from early-stage loss to profitable acquisition. The product demonstrated strong unit economics once trust and recognition were established.
Key Learnings
UX is not limited to screens. The most impactful work here was invisible - flow architecture, input structure, AI pipeline logic. Designing systems and behavior is just as critical as designing interfaces.
In AI products, input design determines output quality. And constraints - speed limits, platform restrictions, no dashboard don't block good design. They force it to be more intentional. Every limitation became a direction.