AI-Assisted Migration Assurance: A Coherent Framework of Cloud Deployment with Security and Reliability

Ghodekar, Apurva Prashant (2025) AI-Assisted Migration Assurance: A Coherent Framework of Cloud Deployment with Security and Reliability. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (995kB) | Preview

Abstract

The automation of cloud migration processes is the new trend, but currently, the vast majority of migration pipelines do not have tools of checking the correctness of infrastructure and its running behaviour post the deployment. Such a loophole leaves organisations vulnerable to configuration fault, security threats, and failure to handle reliability failures, which standard CI/CD pipelines cannot identify. The project creates a single, five-layer cloud migration assurance system comprising of Infrastructure-as-Code provisioning, policy-as-Code validation with Open Policy Agent (OPA), AI-assisted telemetry analysis with the help of LSTM and Autoencoder models, and automated decision-making in a CI/CD pipeline. The system provisions AWS resources with the help of Terraform, implements serverless applications with the help of the Serverless Framework, performs verification of configuration policies prior to deployment, and does runtime anomaly detection with the assistance of real-time CloudWatch metrics. It can be evaluated that OPA policies are effective at identifying misconfigurations, including insecure subnets and weak IAM roles, and the machine-learning models can determine the presence of behavioural anomalies with little telemetry. As the dataset size increases, the model accuracy and stability is enhanced greatly so that the latency spikes and performance deviations can be detected reliably. The findings indicate that policy validation with AI-based runtime analysis is more assuring as compared to the current migration processes. It is a developer-friendly framework that provides a scalable, automated and reproducible framework that will increase the reliability of cloud migrations, correctness and safety, and is a groundbreaking and practical contribution to the research on DevOps and cloud engineering.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Heeney, Sean UNSPECIFIED
Subjects:	Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > QA Mathematics > Computer software > Computer Security T Technology > T Technology (General) > Information Technology > Computer software > Computer Security
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	31 Aug 2026 14:35
Last Modified:	31 Aug 2026 14:35
URI:	https://norma.ncirl.ie/id/eprint/9705

Actions (login required)

View Item