Distributed systems are notoriously difficult to build and manage, with a host of challenges related to transactions, state management, availability, and more. In this talk, I will share my experiences building a distributed system that is designed to be highly available, resilient to outages and disasters, and able to handle long-running processes and overcome interruptions. I will discuss how I tackled the challenges of implementing transactions and managing the state in a distributed environment, and how I was able to overcome the limitations of traditional approaches to achieving 99.99% uptime. I will also share how I was able to ensure correct error handling and avoid domain leakages, and how I made my distributed processes observable and transparent. This talk will be of interest to anyone looking to build and manage data and processes in distributed systems that are reliable and resilient.