DesignLens is an AI-powered design critique platform developed to explore how intelligent feedback systems can support self-taught junior designers who lack access to expert mentorship. The project was motivated by the challenges faced by learners in low-resource environments, particularly in Bangladesh, where access to professional design guidance is often limited.
Grounded in formative feedback theory (Hattie & Timperley, 2007), self-efficacy theory (Bandura, 1997), and Human-Computer Interaction (HCI) principles, DesignLens provides structured critique across five core visual design dimensions: Typography, Color & Contrast, Visual Hierarchy, Alignment & Composition, and Overall Impact. The platform employs vision-capable large language models and multiple pedagogical critique personas to generate actionable feedback tailored to novice designers.
To evaluate the system, a mixed-methods exploratory study was conducted with 12 junior designers in Bangladesh. Participants used DesignLens to iteratively revise their design work while quantitative and qualitative data were collected to examine perceived usefulness, learning outcomes, and design improvement. Findings suggest that AI-mediated critique may serve as a valuable supplementary feedback mechanism in contexts where access to human mentorship is limited.
This repository contains the complete research thesis, methodology, system design documentation, and evaluation results associated with the DesignLens project.
Key Contributions
Development of an operational AI-powered graphic design critique platform.
Application of HCI and educational theory to AI-assisted creative learning.
Empirical evaluation involving junior designers in Bangladesh.
Exploration of AI as a potential tool for reducing mentorship inequities in low-resource environments.
Contribution to ongoing discussions on AI-supported learning, design education, and equitable access to expertise.
Author
Ibrahim Mursalin
Independent Researcher
Mymensingh, Bangladesh