Production Data for Testing, Safely and Securely

Making the impossible, possible.

If we want to replicate the complexities of our production environment, why not use production data? Getting production data into any environment, while ensuring privacy, is a game changer.

Have you ever had a production problem that was difficult to recreate? Do you want test data that emulates real-world complexities? Are you afraid of privacy considerations? If so, this session will show you how we pull data from anywhere to anywhere, safely and securely. The principles presented today hold true for any domain, but this talk is based upon a Guidewire InsuranceSuite implementation. Insurance is complex. You have accounts, policies, processes, claims, contacts, vendors, and payments. This data resides in hundreds of records from dozens of tables across different databases with complex relationships. Many tools navigate a database, so we were excited at the possibility of simply extracting entities and recreating them in another system. Then reality hit, and the work started. I'll take you on our journey from thought to reality, showing the pitfalls we encountered and the results we achieved. You will learn how we: • extract entities from one system and its related entities in other systems, • safely copy entities to other systems, • address privacy concerns with repeatable anonymization when needed, • implemented system-syncing as required, • ignore database types with database-agnostic capabilities, and • purge entities and orphans to reduce a database's size saving time and money. All this from the simple desire to get production data into a test system. By the end, I hope you'll see how to apply some of these principles to your own systems. It was an incredible journey for my company and me, and I'm excited to share it with you.