Fabrice Deseyn

## Overview When implementing a data platform on Azure, it's easy to focus solely on technical execution—getting things working during a proof of concept. But as your data platform evolves and matures into a data mesh architecture, effective governance becomes critical. This is where PlatformOps comes in. Without proper governance, data platforms quickly become fragmented—with inconsistent deployment practices, security vulnerabilities, duplicated infrastructure code, and compliance gaps. PlatformOps provides the framework to prevent this chaos by establishing standardized processes, automated workflows, and clear ownership boundaries. When governed properly, your platform becomes a reliable, scalable foundation that accelerates development velocity rather than hindering it, ensuring teams can innovate safely while maintaining enterprise-grade quality and security standards. This session explores the importance of PlatformOps for governing your Azure Data Platform, with a strong emphasis on standardization and automation as foundational principles. We'll walk through a real-world use case and architecture demonstrating how to implement these principles when building a data platform on Azure Databricks. --- ## What You'll Learn - **Code & Infrastructure Reusability** - Leveraging Infrastructure-as-Code (IaC) patterns - Building reusable Terraform modules and templates - **CI/CD Pipeline Implementation** - Automated deployment workflows for data platforms - End-to-end pipeline orchestration - **Security & Environment Management** - Establishing safe, secure production environments - Implementing robust development and testing workflows - **Azure DevOps + Terraform** - The powerful combination for PlatformOps excellence - Practical integration patterns and best practices --- ## Real-World Insights We'll share practical lessons learned from actual implementations, including: - **Obstacles encountered** during our journey - **Solutions and workarounds** that proved effective - **Key decisions** that shaped our architecture

Databricks’ Secret Scopes are often the default for managing sensitive values like service credentials and encryption keys in Databricks. But as organizations scale, these mechanisms reveal critical shortcomings. In this session, we’ll break down why Secret Scopes fall short when used for credential storage and encryption key management. You’ll learn how these limitations can compromise data access, auditability, lifecycle management, and encryption key segregation. We’ll walk through two core use cases—managing service credentials and securing encryption keys—and demonstrate why the this tooling might not meet enterprise needs for secure data platforms. This session sets the foundation for rethinking secrets management in a scalable, production-grade Databricks platform setup.