Creating good software relies heavily on modeling, which involves representing the design of software systems using diagrams. The Unified Modeling Language (UML) is a standard method for this, but manually developing these diagrams can be complex, time-consuming, and requires significant expertise.
This project aims to simplify the process by automatically generating UML diagrams, such as class diagrams and use case diagrams, from software requirements. Instead of building a custom machine learning model from scratch, this approach uses a Large Language Model (LLM), like GPT, through an API. The application will act as an LLM agent where users input their software requirements. The system will then create specific prompts to communicate with the LLM, which will generate the relevant UML diagrams based on those requirements.
In essence, this project explores using advanced AI tools (LLMs) to automate the generation of UML models, making the process faster, easier, and more accessible for software developers.