EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Autonomous AI Agents for Oncall Investigation and Incident Resolution

Veeranji Reddy Jakkireddy

International Journal of Scientific Research in Computer Science, Engineering and Information Technology · 22.07.2025 · 11 (4) · 825-840

DOI: 10.32628/cseit251116288 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Scientific Research in Computer Science, Engineering and Information Technology, online erschienen am 22.07.2025. Verfasst von Veeranji Reddy Jakkireddy. Die Arbeit nennt 35 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The modern enterprise applications are supported by cloud-native software systems owing to scalability and elasticity of such platforms, and the capability to run applications on distributed microservices architecture. But growing complexity of such environments has led to increased difficulty in investigating incidents on call, as the operational data is constantly produced through various sources such as application logs, infrastructure metrics, distributed traces, monitoring alerts, and service dependency graph. The conventional solutions in Artificial Intelligence for IT Operations (AIOps) mostly concentrate on anomaly detection and root cause analysis individually and need many manual interventions during the incident investigation and resolution processes. Such inefficiencies increase the Mean Time to Detect (MTTD) and Mean Time to Resolve (MTTR), incur high operational costs, and increase service downtimes. In order to mitigate these issues, this paper presents an Autonomous Intelligent Incident Resolution Framework (AIIRF). The multiple AI agents that work in collaboration with each other analyze multi-source observability data and prioritize incidents based on the data, determine probable root causes of incidents, suggest possible actions, and automate the remediation process workflow with the enhancement of decision-making skills with the help of historical operational data. The proposed framework is analyzed with the help of heterogeneous cloud-native observability data that include application logs, infrastructure metrics, distributed traces, monitoring alerts, Kubernetes events, and historical incidents in the enterprise microservices environment. The experimental results show that AIIRF obtains incident detection accuracy of 99.61%, root cause identification accuracy of 99.42%, incident resolution success rate of 99.58%, automation accuracy of 98.73%, and operational performance of 99.51% while lowering the Mean Time to Detect down to 10.9 seconds and Mean Time to Resolve to 13.1 minutes. Comparative analysis with the ServiceRank framework and Context-aware framework shows that AIIRF provides better scalability, intelligent reasoning, operational resilience, and autonomous incident management. It can be concluded that the proposed framework serves as an efficient and scalable solution for next-generation AI-based on-call investigation and incident automation.

Zitieren

APA

Jakkireddy, V. R. (2025). Autonomous AI Agents for Oncall Investigation and Incident Resolution. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 11(4), 825-840. https://doi.org/10.32628/cseit251116288

Vancouver

Jakkireddy VR. Autonomous AI Agents for Oncall Investigation and Incident Resolution. International Journal of Scientific Research in Computer Science, Engineering and Information Technology. 2025;11(4):825-840. doi:10.32628/cseit251116288

BibTeX
@article{Jakkireddy2025,
  author = {Jakkireddy, Veeranji Reddy},
  title = {Autonomous AI Agents for Oncall Investigation and Incident Resolution},
  journal = {International Journal of Scientific Research in Computer Science, Engineering and Information Technology},
  year = {2025},
  volume = {11},
  number = {4},
  pages = {825--840},
  doi = {10.32628/cseit251116288}
}
Kostenlos starten →