Showcase: Cognitive UX Co-Creator — Blending Psychology & AI for Better UX

Google AI Studio
1

​Hi everyone! :waving_hand:

​As a psychology student, I’ve always wondered: Why do most UX tools only focus on visuals, while completely ignoring the user’s mental bandwidth and emotional safety?

To solve this, I built Cognitive UX Co-Creator using Google AI Studio and Gemini API. It’s a workspace prototype designed to help creators spot and fix hidden UX friction through the lens of cognitive psychology.

What problems does it solve?

​:brain: Reduces Cognitive Load: Cuts down mental friction during onboarding and user flows.

​:shield: Ensures Emotional Safety: Helps build trust and reduces user uncertainty in critical moments.

​:triangular_ruler: Brings Science to Design: Evaluates interfaces based on real psychological and decision-making principles.

​:link: Check out the Demo: guided-reflection-ai-cognitive-ux-co-creator.ai.studio

:open_file_folder: GitHub: GitHub - zahrakhorram2314-del/cognitive-ux-cocreator: An AI-powered UX research agent built in Google AI Studio to evaluate cognitive friction, decision load, and accessibility in digital products.5 · GitHub

​Quick question: When you’re building or using AI tools, what’s the biggest mental roadblock you usually face?

I’d love to hear your thoughts and learn from your feedback! :sparkles:

2

Amazing!

How did you accomplished building this app?

3

Thank you so much! :blush:

​I built Cognitive UX Co-Creator by combining Google AI Studio {Gemini API} with core principles from Cognitive Psychology and HCI.

​Here is a quick overview of how I put it together:

​Psychological Frameworks: I mapped out System Instructions grounded in Cognitive Load Theory and emotional safety models to detect mental friction in user flows.

​Google AI Studio & Gemini API: I designed system prompts with specialized evaluation modes {like UX Research, A/B Comparative mode, and Empathetic analysis} to generate structured, actionable redesigns.

​Open Architecture: I shared the complete prompt structure {system_instructions.md} on my GitHub repository so anyone can review or test the underlying logic.

​Are you currently building something with Google AI Studio or exploring UX tools?

I’d love to learn more about your work!