3rd International Conference on Business Intelligence and Data Analytics (BIDA 2026)

Organized by RV Institute of Management (RVIM), Bangalore, India

 Technically Sponsored by Soft Computing Research Society

April 11-12, 2026The after-conference proceeding of the BIDA 2026 will be published in Scopus Indexed Springer Book Series "Smart Innovation, Systems and Technologies"

The after-conference proceeding of the BIDA 2026 will be published in Scopus Indexed Springer Book Series "Smart Innovation, Systems and Technologies"

Sauhard Bhatt

Optimizing Cloud Data Delivery with Selective Deployment Patterns

Abstract:

Modern cloud data platforms have enabled rapid development of data pipelines, but deployment models have not kept pace. In many cloud-based data integration systems, deployments use a full-state promotion approach, in which the entire platform configuration is released as a single unit. While effective for simpler environments, this model introduces significant challenges in large-scale, multi-team scenarios—particularly when features are developed and validated independently.

This talk explores the limitations of full-state deployment in cloud data platforms and presents a practical, engineering-driven approach to enabling selective deployment. The core problem arises when parallel feature development cycles are misaligned: one feature may be production-ready while another is still undergoing validation, yet traditional deployment mechanisms force both to be released together. This lack of deployment granularity increases production risk, delays releases, and creates operational bottlenecks.

To address this, we introduce a selective packaging and deployment pattern that decouples development intent from platform state. By introducing a manifest-driven workflow, dependency-aware artifact selection, and a staged export mechanism, teams can promote only the intended components while preserving the integrity of the overall system. The approach integrates seamlessly into CI/CD pipelines and leverages incremental deployment strategies to ensure safe and controlled releases.

The session will walk through real-world implementation patterns, including dependency resolution, separation of application and infrastructure concerns, and automation of validation and deployment workflows. Key lessons learned—such as the importance of dependency completeness, governance boundaries, and deployment safety—will also be discussed.

By the end of this talk, attendees will understand how to transition from monolithic deployment models to precision-driven release strategies, enabling faster, safer, and more scalable data platform operations in modern cloud environments.

Profile:

Sauhard Bhatt is an accomplished Cloud and DevOps engineering leader with over 19 years of experience architecting, modernizing, and governing enterprise-scale hybrid platforms across healthcare, financial services, and highly regulated environments. He currently leads large-scale cloud transformation initiatives, shaping organizational platform strategy, operational excellence, and security posture. Core expertise spans Azure governance, DevOps automation, identity and access modernization, and middleware transformation across WebLogic, WebSphere, container, and Kubernetes-based platforms.

In his current role, he provides technical and architectural leadership for enterprise-wide initiatives, including Azure Arc–based hybrid governance, standardized CI/CD platforms using GitHub Actions, infrastructure-as-code adoption, and adoption of AI-assisted engineering practices by integrating Microsoft Copilot into development and DevOps workflows. In parallel, he researches and evaluates Internal Developer Platform (IDP) capabilities, focusing on self-service provisioning and failure-proof automation to accelerate delivery while maintaining governance and security controls.