Describe a scenario where you had to optimize an algorithm to run in a low-latency environment at NDA. What profiling tools did you use?
