Kimmel, RobinDhungana, DeepakLambers, LeenBonorden, LeifHenning, Sören2024-02-142024-02-142024https://dl.gi.de/handle/20.500.12116/43511This work examines the potential of blending traditional programming methods with artificial intelligence, specifically large language models (LLMs), to automate the creation of web applications. The primary focus is on defining the necessary software architecture and components to transform diverse inputs, such as natural language and Unified Modeling Language (UML) notations, into functional web applications. The core concept involves a software agent built around a Large Language Model, equipped with tools to autonomously address tasks. While demonstrating promise, this approach exhibits certain limitations that demand further exploration and refinement.enLarge Language ModelsText/Conference Paper10.18420/sw2024-ws_14