Role
Timeline
2 weeks
Year
2024
Type
Overview
NBULA is a low-code platform that democratizes AI model building—enabling non-technical users to create, train, and deploy machine learning models without coding expertise. The product existed only as a concept, and the team needed a complete MVP design within two weeks for investor pitch.
As the solo designer, I conducted research with AI/ML developers, designed the entire user experience from authentication to model deployment, and delivered a functional prototype that secured successful funding.
Key Challenges
Design complete platform from concept in extremely tight 2-week timeline
Simplify complex AI/ML workflows for non-technical users
Create intelligent prompt system that works across any industry or use case
Balance powerful customization with approachable simplicity
Design Strategy
Smart Prompt System: Designed a fixed-structure prompt requiring only 3 user inputs (industry, segment, description) that intelligently generates appropriate AI models—eliminating technical complexity while maintaining flexibility.
Template-First Approach: Created dashboard with pre-built templates (retail forecasting, churn prediction, image classification) allowing users to start immediately with proven use cases.
Visual Workflow Builder: Designed drag-and-connect interface where users see their AI pipeline, compare models in real-time, customize training parameters, and deploy—all without writing code.
Research-Driven Design: Collaborated with AI/ML developers to understand core technical constraints, then translated complex model-building processes into intuitive visual workflows.
Impact
Delivered complete MVP design in 2 weeks that secured funding. The prototype enabled the team to demonstrate their vision to investors and move forward with development—transforming AI model building from a technical barrier into an accessible, visual workflow.
