BRD-to-Jira Automation Engine

Summary

Reduced time from business requirements to engineering-ready tickets by building a semi-automated BRD-to-Jira pipeline. Designed structured markdown inputs and AI agents to generate epics, stories, and acceptance criteria with high consistency. This improved delivery velocity and reduced ambiguity between product and engineering teams.

The problem

Product requirements lived in inconsistent documents and required manual translation into Jira tickets, creating delays, ambiguity, and uneven quality across teams.

What I did

• Designed a structured markdown-based BRD format to standardize inputs → reduced variability in requirements quality
• Built an AI-assisted workflow using Amazon Kiro + LLM APIs to generate epics, stories, and acceptance criteria → cut manual ticket creation time
• Implemented a semi-automated review layer for edge cases → balanced speed with accuracy

Outcome

Reduced time from requirements to Jira-ready tickets from hours to minutes while improving consistency across teams. Established a scalable foundation for AI-assisted product workflows.