Design Decode: AI-powered Design-Led Image Analysis Tool

Project Description: Design Decode is a web-based application developed to support design critique and reflection through multimodal AI-assisted analysis. Rooted in critical pedagogy and cross-disciplinary design theory, the tool enables users to upload visual artefacts—such as graphic compositions, user interfaces, or speculative design outputs—and receive structured analytical feedback across six defined interpretive frameworks.

The six modes of analysis—Visual Form, User Flow, Symbol Logic, Production Logic, Cultural Lens, and Suggest Critical Questions—draw upon methodologies from visual culture studies, interaction design, semiotics, systemic critique, and political theory. These categories are not evaluative checklists but hermeneutic tools. They allow for situated critique and context-aware interpretation, encouraging designers to attend to the layers of meaning embedded within their artefacts. The analytical output is generated using Google’s Gemini 1.5 Flash multimodal model, which responds to both image data (base64-encoded) and (optional) image context statements submitted alongside the image.

Designed with accessibility and clarity in mind, the tool employs a modular interface built using HTML5, Tailwind CSS, and vanilla JavaScript. It features a responsive canvas-based image viewer, method selection interface, accordion-style result display, and export options in plaintext and Markdown formats. The system foregrounds interpretability and avoids normative scoring systems or reductive feedback mechanisms. Instead, it promotes a non-judgmental, position-taking mode of analysis—assertive in tone but grounded in design literacy rather than aesthetic opinion.

The prompt structure embedded within the application has been carefully tuned to resist generic feedback, automated approval, or moralistic drift. The responses are direct, critical, and pedagogically usable—designed to provoke reflection in design classrooms, portfolio settings, and self-led critique.

While the tool can serve individual practitioners, its primary intended context is higher design education, particularly in courses dealing with communication design, interaction design, or critical design methods. By disaggregating visual analysis into layered interpretive axes, Design Decode foregrounds the interrelationship between form, function, message, context, and ethics—thus resisting the flattening effects of standardised critique.

As a research and pedagogical intervention, Design Decode poses questions around how generative AI can be re-purposed not for creation but for interpretation. It reframes the role of AI in design processes—not as a speculative co-creator but as an attentive, repeatable, and rigorously scoped reader. This approach aligns with emerging calls for critical and reflexive uses of AI in education, where the aim is not automation of creativity but augmentation of interpretive depth.

In future iterations, the project seeks to incorporate multilingual support, localised critique models for region-specific visual languages, and plugins for integration into Learning Management Systems (LMS). Moreover, data from anonymised usage may be studied to understand how novice designers engage with critique structures, how AI framing affects visual literacy, and what kinds of questions students learn to ask when supported by multimodal feedback.

Ultimately, Design Decode challenges the instrumentalisation of AI in design practice by repositioning it as a critical reader, one capable of parsing nuance, context, and implication across varied visual formats. It is, to date, the most rigorously conceived AI-powered design-led image analysis tool available, uniquely tailored for educators, researchers, and critically engaged practitioners.

Developed by Rahul Bhattacharya | Powered by dotai | Prototype for Testing

Design Decode is the most advanced design-led visual analysis tool available online. It combines formal, cultural, ethical, and systems-based interpretation in one interface. Created to support critical, context-aware design thinking, this tool is currently in testing. Your feedback shapes its evolution.