Video by The Linux Foundation via YouTube
Sponsored by Harness
AI coding tools are helping engineering teams create and ship software at unprecedented speed. But our research from 700 enterprise practitioners and engineering leaders reveals a troubling counterweight: 71% say releases or production issues require developers to work evenings or weekends at least weekly, while more than half report frequent or constant burnout.
The surprising finding is that developers do not blame AI. The teams experiencing the most burnout are also among its most enthusiastic users. The real problem is the widening mismatch between AI-scale code generation and delivery systems still dependent on manual testing, fragmented tools, human-operated releases, and reactive incident response.
In this session, attendees will learn:
Why after-hours work should be treated as a software-delivery metric
How high deployment velocity can conceal weak stability and growing rework
Why AI adoption amplifies the strengths and weaknesses of the underlying delivery system
Which capabilities help reduce repetitive toil, deployment risk, and incident burden
How golden paths, automated testing and governance, progressive delivery, verification, rollback, and resilience testing can make AI-scale delivery sustainable
The session will close with a practical demonstration of how the Harness Software Delivery Agent helps delivery systems validate, govern, deploy, and recover without requiring people to absorb every failure manually.
Demo: https://www.harness.io/demo/software-delivery-agent