RANCANG BANGUN SISTEM GENERATOR ANTARMUKA PENGGUNA BERBASIS ATOMIC DESIGN

Mufid Bahaudin Nugroho, NIM.: 22106050021 (2026) RANCANG BANGUN SISTEM GENERATOR ANTARMUKA PENGGUNA BERBASIS ATOMIC DESIGN. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (RANCANG BANGUN SISTEM GENERATOR ANTARMUKA PENGGUNA BERBASIS ATOMIC DESIGN) 22106050021_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (5MB) | Preview
	Text (RANCANG BANGUN SISTEM GENERATOR ANTARMUKA PENGGUNA BERBASIS ATOMIC DESIGN) 22106050021_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (14MB) | Request a copy

Abstract

The rapid growth of the modern software industry demands high-quality, consistent, and maintainable user interfaces (UI). Although Generative Artificial Intelligence (AI) technology has opened opportunities for automating UI code generation, the resulting code is often unstructured, difficult to maintain, and inconsistent with standardized component development patterns. This research aims to design and develop an Atomic Design-based UI generator system utilizing a Large Language Model (LLM) as its code generation engine. The system was developed using the Next.js framework, integrating frontend and backend layers within a unified framework, and utilizing the Google Gemini API as the LLM model. The Atomic Design approach was implemented through a layered prompt engineering technique that explicitly embeds the Atom, Molecule, Organism, Template, and Page hierarchy instructions into every prompt. The system supports three input modes Standard Prompt, Template Builder, and Structured Form to accommodate various user needs ranging from creative exploration to controlled academic testing. Evaluation was conducted through output testing and black-box testing across 12 functional scenarios. The results demonstrate that the system successfully generates UI components automatically in both HTML/Tailwind CSS and React JSX formats, with consistent and modular structures aligned with Atomic Design standards. All black-box testing scenarios achieved a Passed status. The layered prompt engineering mechanism proved effective in guiding the AI model to produce interface code that is consistent, modular, and compliant with Atomic Design principles.

Share this knowledge with your friends :

Actions (login required)

View Item