Requirements are All You Need: From Requirements to Code with LLMs (2406.10101v2)

Abstract: The pervasive use of textual formats in the documentation of software requirements presents a great opportunity for applying LLMs to software engineering tasks. High-quality software requirements not only enhance the manual software development process but also position organizations to fully harness the potential of the emerging LLMs technology. This paper introduces a tailored LLM for automating the generation of code snippets from well-structured requirements documents. This LLM is augmented with knowledge, heuristics, and instructions that are pertinent to the software development process, requirements analysis, object-oriented design, and test-driven development, effectively emulating the expertise of a seasoned software engineer. We introduce a "Progressive Prompting" method that allows software engineers to engage with this LLM in a stepwise manner. Through this approach, the LLM incrementally tackles software development tasks by interpreting the provided requirements to extract functional requirements, using these to create object-oriented models, and subsequently generating unit tests and code based on the object-oriented designs. We demonstrate the LLM's proficiency in comprehending intricate user requirements and producing robust design and code solutions through a case study focused on the development of a web project. This study underscores the potential of integrating LLMs into the software development workflow to significantly enhance both efficiency and quality. The tailored LLM is available at https://chat.openai.com/g/g-bahoiKzkB-software-engineer-gpt.

The paper "Requirements are All You Need: From Requirements to Code with LLMs" explores the integration of LLMs into software engineering, specifically focusing on the automation of code generation from textual software requirements. By capitalizing on the widespread use of text-based documentation for software requirements, the authors develop a tailored LLM that assists in various stages of the software development lifecycle.

Key Contributions:

1. Tailored LLM for Automation: The paper presents an LLM customized for generating code snippets directly from well-structured requirements documents. This model is augmented with domain-specific knowledge and practical heuristics related to software development processes, such as requirements analysis, object-oriented design, and test-driven development.
2. Progressive Prompting Method: The authors introduce a "Progressive Prompting" technique that facilitates interaction between software engineers and the LLM. This approach allows the model to incrementally address development tasks. It begins by interpreting given requirements to identify functional specifications, then progresses to create object-oriented models, followed by generating unit tests and implementing the corresponding code.
3. Emulation of Expertise: The LLM is designed to replicate the expertise of a seasoned software engineer. By embedding relevant instructions and heuristics, the model effectively bridges the gap between human expertise and automated software generation.
4. Case Study and Demonstration: To validate the model’s capabilities, the paper includes a case paper centered around a web development project. The LLM demonstrated its proficiency by accurately understanding complex requirements and generating robust design and code solutions.
5. Impact on Software Development Workflow: The research highlights the potential benefits of incorporating LLMs into the software development process. By automating code generation, organizations can enhance efficiency and improve the quality of software solutions.

Overall, the paper suggests that using LLMs in software engineering can streamline the process from requirements gathering to code creation, significantly boosting productivity and quality within the software development lifecycle.