Check your understanding: estimating concurrency

A simple way to estimate needed concurrency is:

concurrency ≈ requests_per_second * average_duration_seconds

If your function averages 200 ms per request and receives 50 requests/second, what concurrency do you need (approximately)?