AI-Driven "Immunological" Drift Detection in Serverless Workflows

Abstract

This research considers immunologically inspired AI models for serverless environment detection and specifically, for detecting behavioural drift in AWS Lambda environments. The model uses artificial immune systems and federated learning, i.e., precision latency independent of cold start and dependency change anomalies are identified with high precision and low latency resulting in huge improvement in workflow reliability, SLA adherence and real time diagnostics.