Designing Cloud-Native Reference Architectures for Enterprise-Scale AI/ML Platforms

Santosh Pashikanti

doi:10.71097/IJAIDR.v16.i2.1638

Designing Cloud-Native Reference Architectures for Enterprise-Scale AI/ML Platforms

Author(s)	Santosh Pashikanti
Country	United States
Abstract	Enterprises are racing to operationalize AI/ML at scale, but many initiatives stall in a maze of bespoke pipelines, siloed tools, and inconsistent deployment models. In my experience designing multi-cloud platforms, the missing piece is often a repeatable, cloud-native reference architecture that standardizes the end-to-end AI/ML lifecycle from data ingestion and feature management to training, deployment, and runtime governance. This paper proposes a set of practical, vendor-agnostic reference architectures for enterprise AI/ML platforms built on Kubernetes, microservices, and managed cloud AI services. I first examine related work and the current industry landscape around cloud-native AI, MLOps, and feature stores. I then derive system requirements and design principles for platform blueprints that support heterogeneous workloads (batch, streaming, real-time, generative AI), multi-cloud deployment, and strict security/compliance needs. The paper presents a modular architecture with clearly defined domains Ingestion, Feature Store, Training, Serving, and Cross-Cutting Services mapped to Kubernetes-native and managed services across AWS, GCP, and Azure. A detailed case study of a global enterprise AI platform illustrates how these blueprints can be implemented using Kubernetes, service mesh, feature stores, workflow engines, and managed AI offerings. I discuss evaluation criteria such as model iteration lead time, infrastructure utilization, reliability, and portability, and present indicative results from real-world deployments. The paper concludes with a discussion of trade-offs, limitations, and practical guidance for adopting these reference architectures as organizational standards for AI/ML platform engineering.
Keywords	Cloud-native computing, Kubernetes, MLOps, AI platforms, reference architecture, microservices, feature store, multi-cloud, model serving, data ingestion, managed AI services.
Field	Engineering
Published In	Volume 16, Issue 2, July-December 2025
Published On	2025-08-08
Cite This	Designing Cloud-Native Reference Architectures for Enterprise-Scale AI/ML Platforms - Santosh Pashikanti - IJAIDR Volume 16, Issue 2, July-December 2025. DOI 10.71097/IJAIDR.v16.i2.1638
DOI	https://doi.org/10.71097/IJAIDR.v16.i2.1638
Short DOI	https://doi.org/hbf77k

View / Download PDF File

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJAIDR DOI prefix is
10.71097/IJAIDR

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 16 Isu 2 Cover Page Vol 16 Isu 1 Cover Page Vol 15 Isu 2

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJAIDR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijaidr.com

Journal of Advances in Developmental Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Designing Cloud-Native Reference Architectures for Enterprise-Scale AI/ML Platforms

Share this