2024
Role
Research, UX strategy & validation
Timeline
~6 months
Team
Key impact
Reduced user confusion
Clarified system behavior and interaction expectations
↑
Improved prompt success
Users reached expected outcomes with fewer iterations
↑
Feature discoverability
Core capabilities became easier to find and understand
↑
User-driven decision making
Shifted discussions from assumptions to validated insights
Overview of the prompt-based no-code platform and its conversational app-building workflow
Context
The project focused on improving the usability of an AI-powered no-code platform that allowed users to build applications through prompt-based interactions.
At the time, the company was still in its early stages, initially releasing the product to friends and family before expanding to beta testers and public users.
As adoption grew, I advocated for introducing UX research into the product process to better understand how non-technical users interacted with the platform and where usability friction emerged.
From that point on, I conducted ongoing usability testing and user interviews throughout the product’s growth, helping guide future improvements based on real user behavior.
Understanding user behavior
Usability sessions revealed a recurring pattern: the platform was designed around a technical mental model, while many users approached it with little understanding of how AI-driven systems behaved.
This gap created friction throughout the experience, especially during prompt-based interactions and moments where the system lacked clear guidance or feedback.
Usability testing session with new users
Key takeaways from research
Users struggled to predict system behavior
Technical terminology created onboarding friction
Feedback visibility impacted confidence
Prompt interactions felt too open-ended
Defining the design direction
The research revealed that many usability issues were not caused by isolated interface problems, but by a mismatch between how the system was designed internally and how users expected AI-driven interactions to behave.
Non-technical users often approached the platform without a clear understanding of:
How prompts influenced outcomes
What the system was doing
How to recover when results didn’t match expectations
Rather than focusing only on adding new features, the design direction prioritized:
Improving clarity
Reducing ambiguity
Strengthening feedback visibility
Supporting users through prompt-based interactions
The goal became helping users build a clearer mental model of how the platform behaved throughout the app creation process.
Mental model mismatch
Product improvements
Rather than redesigning the platform from scratch, my role focused on translating usability insights into improvements that could better support non-technical users while fitting the realities of an evolving startup product.
The recommendations focused on reducing ambiguity, improving system communication, and helping users better understand how to interact with AI-driven workflows.
Clarifying terminology
Problem
Many interface labels and concepts reflected internal technical language that was unfamiliar to non-technical users.
Improvement
I worked with the team to simplify terminology and restructure certain labels to better align with how users naturally understood tasks and workflows.
Why it mattered
Reducing technical ambiguity made the platform easier to approach and lowered onboarding friction during early interactions.
Replacing technical terminology reduced onboarding friction for non-technical users
Structuring feedback
Problem
Users often struggled to understand whether actions had succeeded, what the system was doing, or what should happen next.
Improvement
I collaborated with the team to improve confirmation patterns, messaging, and feedback visibility throughout key interactions.
Why it mattered
Clearer system feedback reduced uncertainty and improved user confidence during app creation tasks.
Users could review generated changes, understand what was modified, and recover previous versions when needed
Supporting prompt interactions
Problem
Prompt-based interactions gave users flexibility, but many struggled with how to structure prompts or recover when results didn’t match expectations.
Improvement
We explored ways to support users through examples, contextual guidance, and more structured prompt interactions without removing creative flexibility.
Why it mattered
Additional guidance helped users interact with the platform more confidently while preserving the exploratory nature of AI-driven workflows.
Prompt suggestions reduced blank-state friction during app creation
Influencing product decisions
Problem
As a fast-moving startup product, many decisions were initially driven by internal assumptions rather than direct user behavior
Improvement
I introduced recurring usability testing and shared findings directly with the multidisciplinary team to help guide prioritization and product discussions
Why it mattered
Continuous user feedback shifted conversations from assumptions toward validated insights, helping the team make more informed product decisions over time
Iteration & collaboration
Usability testing became a continuous part of the product workflow as the platform evolved.
Rather than relying on formal metrics, I evaluated improvements by observing whether users could move through flows without getting stuck at the same recurring friction points identified in earlier sessions.
As users became more comfortable completing tested interactions, new areas of the experience were introduced into testing cycles. These sessions also helped uncover broader feedback, feature interests, and evolving user expectations.
Insights were regularly shared with the multidisciplinary team, helping connect user behavior to prioritization and product discussions throughout the platform’s growth.
Reflection
This project reinforced the importance of introducing continuous user validation early in fast-moving AI product environments.
Observing how users interpreted prompts, feedback, and system behavior revealed recurring gaps between internal technical assumptions and real user expectations.
It also highlighted how usability in AI products goes beyond interface design — requiring systems to help users build confidence and clear mental models throughout interactions.
