Throughput
How much work a system completes per unit of time: requests answered, records processed, or data moved per second.
What it tells you about a developer
Throughput fluency signals capacity thinking. This developer has estimated what a system can handle, load-tested it to check, and planned for the day traffic doubles. That work exists at companies with heavy request volume, so it usually indicates scale in their background. Engineers who have done this work tend to remember their requests per second, a number that sticks because the work behind it was real.
Throughput and latency measure different things. Latency is how fast one request finishes, throughput is how many finish per second. A system can be strong at one and weak at the other, the way a freight train carries enormous loads slowly. Job descriptions that say high-throughput mean the system handles very large request volumes, and capacity planning is the work of predicting how much it can take.