Copilot after 18 months on a large monorepo - real experience

We've been running GitHub Copilot across a ~500k line monorepo for a year and a half. Multiple services, shared libraries, the works.

Here's the real take:
* It's not about writing whole functions. It's about saving 20-30% of keystrokes on boilerplate and standard patterns.
* Biggest win: filling out test cases and repetitive configs (Dockerfiles, CI steps, YAML).
* It struggles with our internal domain logic unless the file has strong, clear types.
* Context window limits mean it often misses relevant patterns from other parts of the repo.

Bottom line: It's a productivity bump, not magic. You still need to know exactly what you're doing. It pays for itself by cutting down the tedious stuff, but you can't let it drive.