Developer builds custom LLM pipeline to auto-groom 500 Jira tickets

A developer has created a custom pipeline to automatically process and organize Jira tickets using machine learning and LLMs, addressing a gap in Atlassian's native tools. The process involves NLP preprocessing, TF-IDF vectorization, and K-Means clustering to group tickets thematically and detect duplicates. Subsequently, Gemini 2.5 Flash is used with selective RAG grounding to generate enriched cluster names, identify age-aware duplicate insights, and produce an executive summary. AI

IMPACT This approach demonstrates how custom LLM pipelines can augment existing tools to perform complex batch analytical tasks beyond their native capabilities.

RANK_REASON Developer describes a custom tool/pipeline built using ML and LLMs for a specific task.

Read on dev.to — LLM tag →

• Apache ZooKeeper
• arXiv
• Atlassian
• Atlassian Intelligence
• Gemini
• Gemini 2.5 Flash
• GenAI-Enabled Backlog Grooming in Agile Software Projects
• GPT-4o
• Jira
• k-means clustering
• retrieval-augmented generation
• Rovo Agents
• tf–idf

• product
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →