Selected work
From Service Desk to Product Intelligence
I designed an AI-assisted monthly workflow that turned recurring Service Desk activity into structured decision support for Service Desk Operations, Product, and Engineering.
The challenge
Support requests have to be handled one at a time, but a monthly population of tickets also contains signals: recurring defects, usability friction, possible product needs, aging work, and resolution patterns. The challenge was to make those signals visible without turning ticket frequency into an automatic product or engineering decision.
The workflow
On a monthly cadence, Jira Automation and JQL selected the relevant Service Desk requests. An AI agent in the Atlassian environment categorized requests, grouped recurring themes, identified patterns, and prepared an executive synthesis. The combined analysis and operational view were then published to Confluence for asynchronous review.
One source, three decision contexts
For Service Desk Operations, the report surfaced SLA performance, aging tickets, and resolution patterns. For Product, it provided signals about recurring requests, user-experience friction, and potential feature opportunities. For Engineering, it made recurring defects and product behavior easier to discuss as part of prioritization. The workflow informed decisions; it did not make them autonomously.
My role and value
I designed and implemented the workflow: the reporting use case, JQL-based selection, automation, AI-assisted categories and insights, Confluence output, and the human-in-the-loop operating model. It created a recurring mechanism for connecting support activity with product and engineering feedback loops, while avoiding unsupported public claims about business impact.