Epic Systems

Good pay in a low COL area. People there are nice. Subsidized food, your own office (good and bad). You can set your boundaries and choose how hard you work. Your compensation will reflect your decision, and they are fine with what you choose.

Company cares about your growth (within the company), so you have some freedom on what internal projects and responsibilities you want. You can choose more leadership versus technical as well.

Good stepping stone to FAANG or other jobs, even if the tech stack leaves much to be desired (FAANG doesn't care, since they are going to train you anyways).

In my opinion, if you can navigate Epic and its disaster of tech, everything else will come easy.

The platform is a behemoth tied together by a lot of super glue. Most of the business logic is written in a language called Mumps, which is incredibly limited in features.

You can forget about data structures/classes and anything related to modern software development. For the most part, you're mostly doing CRUD with some extra steps in the middle. Be prepared to deal with a lot of spaghetti code. The inability to abstract business logic in a decent way is already hurting the company's ability to build features in a timely manner (Hyperspace Web was years behind schedule).

There's no widespread automated unit testing or integration testing, but I hear they're trying to make it more standard. This reinforces the spaghetti and makes refactoring impossible and discouraged. Testing is done by QA type roles.

You do get a taste of modern programming with TypeScript, but that's limited to FrontEnd development.

However, it seems like the way the internal metrics are set up, Epic somehow prefers the company spend 100 hours each on 5 features (for a total of 500 hours) compared to 200 hours spent on proper architecture and then 30 hours spent each on the 5 features (350 hours).

Respect best practices when it comes to software engineering, code style, and architecture. Worry less about billable hours. Hire better statisticians and fix your internal metrics, because you are interpreting the wrong lessons from the data.