
Jira tickets were never built for AI coding agents, until now
84% of developers use AI coding tools. 41% of all code pushed to production is AI generated. Code generation, test writing, PR reviews, the downstream layer moved fast. The upstream layer did not move at all. IDC's 2024 survey found engineers spend only 16% of their time on actual application development. The rest goes to operational tasks, CI/CD, security, and coordination. Microsoft Research found something adjacent. Developers want to spend more time architecting and designing systems. They consistently cannot because that work depends on tribal knowledge held by very few people. The Jira ticket sits at the center of this problem. A product manager creates an epic. Two sentences, a vague acceptance criteria, maybe a Figma link. That ticket becomes the input for the entire downstream workflow. In a human workflow, context gaps got filled through conversation. The developer walked over, asked the senior engineer what the ticket meant, clarified service boundaries, and proceeded. AI co
Continue reading on Dev.to
Opens in a new tab




