Graph-Based AI System for Real-Time Detection and Rollback of Performance Regressions in Software Deployments
DOI:
https://doi.org/10.31224/4873Keywords:
Graph Neural Networks, microservices, Performance Regression, Anomaly Detection, Automated Rollback, CI/CD Pipeline Software Reliability, Deployment Monitoring, Distributed Tracing, Scalable AIAbstract
This paper presents a Graph-Based AI System leveraging Graph Neural Networks (GNNs) for real-time detection and rollback of performance regressions in software deployments. By modeling microservices and their interdependencies as dynamic temporal graphs, this system detects anomalies in real time and initiates automated rollback procedures. The integration with CI/CD pipelines and orchestration platforms like Kubernetes ensures scalability and rapid recovery, making it a transformative approach for maintaining software reliability.
Downloads
Download data is not yet available.
Additional Files
Posted
2025-07-17
License
Copyright (c) 2025 Piyushkumar Patel
This work is licensed under a Creative Commons Attribution 4.0 International License.
