International Journal of Scientific Research and Engineering Development

Software systems fail and if you've spent any time in operations, you already know that most of those failures didn't come out of nowhere. Someone pushed a risky commit. A deployment went out at the wrong time. A configuration change slipped through without a proper review. The signal was there; nobody was watching for it. That is the problem this paper tries to address.
Artificial Intelligence for IT Operations (AIOps) has made genuine progress over the last decade. Modern platforms can parse millions of log lines per second, detect anomalies in metric streams, and correlate alerts across dozens of microservices. But almost all of them share the same blind spot: they watch what the system is doing, not what the people building the system were doing before something went wrong. Developer activity the commits, deployments, pull requests, and rollbacks that precede most production incidents is largely invisible to existing AIOps pipelines.
This paper proposes a framework that closes that gap. We argue that by combining traditional infrastructure telemetry with structured signals drawn directly from developer workflows, it becomes possible to detect incident risk earlier and with greater contextual clarity than infrastructure monitoring alone can provide. We present a layered conceptual architecture for this integration, identify the specific developer signals that carry the most predictive weight based on evidence from the empirical software engineering literature, and discuss the practical challenges and ethical considerations that come with deploying such a system in a real engineering organization.

V.Udhayakumar, Amirtha V,"Incorporating Developer Activity Signals into AIOps for Improved Incident Prediction" International Journal of Scientific Research and Engineering Development, V9(3): Page(1465-1472) May-June 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.