What I like most about Step Functions is how simple AWS becomes when using the tool. A lot of complexity typically comes from working with custom code for things like branching, handling errors or retries, etc., and Step Functions helps strip away or obfuscate a lot of that. I also really like how intuitive the process workflow visuals are. These visual representations really allowed me as the end user to know how the workflow proceeds through each step, and what happens upon a failure of a specific step being executed. This has allowed me to more quickly identify issues and bottlenecks, greatly reducing troubleshooting time.

The language used for designing and defining workflows is JSON-ish, which tends to be pretty bulky. The verbosity can make longer processes difficult to read and update, especially if you're a newer user. If you're not experienced with AWS services, then troubleshooting systems across multiple services or accounts can also be fairly cumbersome.

We needed a better way of chaining together the technology pieces behind our business processes. AWS Step Functions allowed us to seamlessly puzzle-piece together our AWS services that serve these processes, and helped reduce the complexity of error handling and troubleshooting the applications behind them.