A web-based software tool utilizing AI agents and prompt engineering to automate task prioritization and apply diverse prioritization techniques, aimed at enhancing project management within the agile framework is introduced.
Abstract
Large Language Models (LLMs) are revolutionizing Software Engineering (SE) by introducing innovative methods for tasks such as collecting requirements, designing software, generating code, and creating test cases, among others. This article focuses on requirements engineering, typically seen as the initial phase of software development that involves multiple system stakeholders. Despite its key role, the challenge of identifying requirements and satisfying all stakeholders within time and budget constraints remains significant. To address the challenges in requirements engineering, this study introduces a web-based software tool utilizing AI agents and prompt engineering to automate task prioritization and apply diverse prioritization techniques, aimed at enhancing project management within the agile framework. This approach seeks to transform the prioritization of agile requirements, tackling the substantial challenge of meeting stakeholder needs within set time and budget limits. Furthermore, the source code of our developed prototype is available on GitHub, allowing for further experimentation and prioritization of requirements, facilitating research and practical application.
In Agile software development (ASD) projects requirements are incrementally and iteratively defined, with customer needs frequently expressed in User Stories (USs). However, minimal documentation has been identified as a key challenge for Requirements Engineering (RE) in ASD. Including too little information makes tracing and estimating USs more difficult, while including too many details limit the developer in their solution. In addition, development teams mainly rely on information contained in issue tracking systems, rather than speaking with their customer on a regular basis. To summarize, development teams are highly dependent on few artifacts.
This PhD dissertation studies how development teams create and use RE artifacts such as requirements and acceptance criteria, guided by the following main research question: How are Requirements Engineering artifacts used in Agile Software Development?
First, we introduce the RE4SA model as a means to support communication between requirements engineers and software architects, recognizing that requirements and architectural components should be designed in tandem. In practice, however, this alignment is difficult to achieve, often due to a lack of concrete guidance in existing models. The RE4SA model addresses this by expressing requirements as epic stories and USs, which are linked to architectural modules and features, respectively. The model is further instantiated as RE4SA-Agile, which connects common agile artifacts and introduces metrics to measure the alignment and granularity between requirements and architecture. These metrics help identify problematic situations, such as when the granularity of requirements or architectural components is inconsistent with the norm.
Then, we focus on the definition of key concepts in the field of RE. Concepts, such as those in the RE4SA model, are often interpreted in different ways. To clarify fundamental concepts in software engineering, we propose the Concept Definition Review (CDR) method. The CDR method was formalized in a second iteration, in which we defined and compared the terms “non-functional requirement” and “quality requirement”, which revealed the existence of dozens of definitions, many nearly identical, and highlighting the importance of systematic conceptual analysis for effective communication.
We also explored the impact of RE artifacts on efficiency of agile teams, focusing on the use of USs and acceptance criteria by teams. Our empirical studies show that while the quality of USs does not directly correlate with timely completion, the existence of acceptance criteria does improve efficiency; we found evidence for an increase in on-time completion and reduced completion time. Our Canonical Action Research study shows that interventions based on the Quality User Story (QUS) framework can improve the quality of USs, but practitioners sometimes find value in deviating from strict guidelines. This suggests that while guidelines are useful, they must be adaptable to the context and practitioner needs. Finally, the challenge of specifying non-functional requirements (NFRs) is addressed; practitioners expressed a need for support in defining NFRs. Unlike functional requirements, NFRs lack a widely adopted writing format and are notoriously difficult to quantify. A new NFR template is developed and validated, based on requirements from practitioners, incorporating fit criteria to make NFRs more measurable.
Algorithm-Driven Development is introduced, a methodology developed from industrial practice to address recurring challenges in translating requirements into reliable, testable, and maintainable software behavior that provides systematic coverage of functional scenarios from the outset of development.
Philippe Jawish, Pierre Evrard, Alexandre Lemerle et al.· Journal of Systems and Softw...· 0 citations
This research aims to identify and validate key challenges and their solutions within the RE process for open‐source software development (OSSD) and propose best practices to address these challenges.
Fazli Rabi, M. Ilyas, Nasir Rashid et al.· Journal of Software: Evoluti...· 0 citations
A lifecycle-aware framework that integrates quantitative software quality assessment with Large Language Model (LLM)-based code refinement is proposed and the potential of metric-driven LLM feedback for research software quality improvement is demonstrated while highlighting its inherently multi-objective nature.
Nafis Tanveer Islam, N. Soveizi, Yutong Li et al.· 0 citations
Requirements prioritization is a fundamental activity in requirements engineering, as it directly impacts release planning, resource allocation, and project success. In agile software development (ASD), this process becomes more challenging due to evolving requirements, multiple stakeholders, limited scalability, and the high cognitive burden of traditional techniques. This study proposes a hybrid prioritization framework that combines the hierarchical cumulative voting (HCV), used to structure requirements and simplify stakeholder input, with the best-worst method (BWM), used to determine consistent expert weights with fewer pairwise comparisons. The stakeholder and expert rankings were integrated using reciprocal rank fusion (RRF) to produce a unified ranking. The framework was evaluated through a case study of an Organizational Quality Management System (OQMS), which included 17 requirements, 43 stakeholders, and 10 experts. The results showed that the approach was effective and reliable, demonstrating that the proposed framework reduced the effort required for requirements prioritization while maintaining a high level of judgment consistency and producing a reliable prioritization result. Overall, the framework demonstrated effectiveness and scalability for ASD while preserving the perspectives of stakeholders and experts.
Salma A. Huwedi, Salwa M. Elakeili· Al-imad Journal of Humanitie...· 0 citations
Program Increment (PI) Planning is a critical synchronization event in scaled Agile methodologies, yet remains largely informal, with limited analytical rigor or strategic alignment. This paper introduces a formal metamodel and a model-driven framework to support structured PI Planning in large-scale software development. Our metamodel captures essential planning artifacts, including features, teams, iterations, risks, and dependencies, and is enriched with OCL constraints for validation and traceability. We extend the model to align planning with business value streams and introduce stakeholder-specific viewpoints for tailored analysis. Compared to tools like Jira or Mural, our approach provides formal semantics, automated reasoning, and extensibility. This work lays the foundation for a Domain-Specific Language (DSL) for PI Planning, bridging the gap between Agile practice and formal modeling.
Le Program Increment (PI) Planning est un événement de synchronisation essentiel dans les méthodologies d'agilité à l’échelle, mais il reste largement informel, avec une rigueur analytique et un alignement stratégique limités. Cet article présente un métamodèle formel et un cadre fondé sur l’ingénierie dirigée par les modèles (model-driven engineering) pour soutenir un PI Planning structuré dans le développement logiciel à grande échelle.Notre métamodèle capture les artefacts essentiels de planification, tels que les fonctionnalités (features), les équipes, les itérations, les risques et les dépendances, et il est enrichi de contraintes OCL pour la validation et la traçabilité. Nous étendons ce modèle afin d’aligner la planification sur les value streams métier et introduisons des vues spécifiques aux parties prenantes pour une analyse ciblée. Comparée à des outils tels que Jira ou Mural, notre approche fournit une sémantique formelle, un raisonnement automatisé et une extensibilité accrue.Ce travail pose les bases d’un langage spécifique (Domain-Specific Language, DSL) dédié au PI Planning, comblant ainsi le fossé entre la pratique Agile et la modélisation formelle.
F. Somda, D. Guel, Kisito K. Kabore· Revue Africaine de Recherche...· 0 citations
Related blog posts
MIT News · Artificial Intelligence· news.mit.eduJun 3, 2026
The new ChartNet training dataset could improve the accuracy of vision-language models that help analyze business trends or interpret scientific figures.
A new method for surgically removing training examples from a model reveals that as datasets grow, the link between what a model learns and what it produces dissolves.
Microsoft Research Blog· microsoft.comJul 30, 2026
Computer-use AI agents struggle with multi-step workflows like email and customer support. Echoverse trains agents in realistic environments rather than simply providing more training tasks, helping them improve as the tasks, tests, and environments evolve. The post Echoverse: Deep, evolving environments for computer-use agents appeared first on Microsoft Research.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
