Code review is a practice for improving the overall quality of a system by closely examining the source code that makes it up. It is intended to detect and fix bugs as well as to improve the overall architecture.
Review usually involves someone who wasn’t the primary author of the code. They examine the code to understand, critique, and question it. They detect bugs, identify areas of improvement, and verify the assumptions that the primary author may have had.
One way we accomplish immediate code review is through pair programming. Pull requests are another way, which work well especially when the reviewer is in a different office, or otherwise cannot pair.
Just like an author wouldn’t publish a book without editing, we won’t ship software without code review.