GenAI-Enabled Backlog Grooming in Agile Software Projects: An Empirical Study

2025-07-14

AI Summary

In agile development, inefficient backlog prioritization arises from redundancy, obsolescence, and ambiguous task definitions. Method: This study proposes a generative AI–driven automation framework for backlog refinement, integrating vector database indexing with the GPT-4o model to perform semantic embedding, cosine similarity–based deduplication, task merging, and cleanup recommendations; following the design science research paradigm, we developed a Jira-integrated plugin prototype. Contribution/Results: Evaluated on real-world project data, the method achieves 100% detection accuracy for redundant items, reduces task processing time by 45%, and significantly enhances backlog management efficiency, transparency, and team collaboration. This work represents the first systematic integration of generative AI and vector retrieval for backlog optimization, delivering a reproducible technical pathway and empirical validation for AI-augmented agile practices.

Technology Category

• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Search and Optimization: Metareasoning and Metaheuristics
• Data Mining & Knowledge Management: Intelligent Query Processing

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Effective backlog management is critical for ensuring that development teams remain aligned with evolving requirements and stakeholder expectations. However, as product backlogs consistently grow in scale and complexity, they tend to become cluttered with redundant, outdated, or poorly defined tasks, complicating prioritization and decision making processes. This study investigates whether a generative-AI (GenAI) assistant can automate backlog grooming in Agile software projects without sacrificing accuracy or transparency. Through Design Science cycles, we developed a Jira plug-in that embeds backlog issues with the vector database, detects duplicates via cosine similarity, and leverage the GPT-4o model to propose merges, deletions, or new issues. We found that AI-assisted backlog grooming achieved 100 percent precision while reducing the time-to-completion by 45 percent. The findings demonstrated the tool's potential to streamline backlog refinement processes while improving user experiences.

Problem

• Automate backlog grooming in Agile projects using GenAI
• Reduce redundancy and improve task clarity in backlogs
• Enhance prioritization accuracy and decision-making efficiency

Innovation

• GenAI automates backlog grooming in Agile
• Jira plug-in uses vector database and GPT-4o
• AI achieves 100% precision, 45% time savings