Impact Factor (2025): 6.9
DOI Prefix: 10.47001/IRJIET
Vol 9 No 11 (2025): Volume 9, Issue 11, November 2025 | Pages: 434-438
International Research Journal of Innovations in Engineering and Technology
OPEN ACCESS | Research Article | Published Date: 01-12-2025
The boom of artificial intelligence and machine learning applications led to the growing need for robust, repeatable infrastructure management practices. Manual provisioning, which is slow and error-prone, was mostly seen in use until recent times. The immediate proposed solution here leverages Infrastructure as Code with Terraform integrated with Policy-as-Code principles. Details that follow are reusable blueprints of Terraform modules illustrating how core AI infrastructure can be provisioned both on Google Cloud Platform and Amazon Web Services, targeting services like Vertex AI and SageMaker. Quantitative benefits analysis through research that provides empirical evidence in observed acceleration of provisioning time, accompanied by tangible cost reductions. In synthesis, Infrastructure as Code together with Policy-as-Code forms Secure, Efficient, Auditable MLOps Environments that remove teams from manual toil and instead allow a shift left to innovation.
Infrastructure as Code, Terraform, MLOps, Policy-as-Code, Cloud Computing, DevSecOps, Vertex AI, Amazon SageMaker
Vatsal Kishorbhai Mavani. (2025). Terraform Modules for AI Infrastructure: Accelerating GCP/AWS Provisioning with Policy-as-Code. International Research Journal of Innovations in Engineering and Technology - IRJIET, 9(11), 434-438 Article DOI https://doi.org/10.47001/IRJIET/2025.911047
This work is licensed under Creative common Attribution Non Commercial 4.0 Internation Licence