Unsure of what the limitations of the program are.

[Pragmer]

It is my understanding that comparable programs to this one can be set to not just recognize a word when it is said, but recognize when it is present inside another word. To give a few examples: the words "Ice" and "nICE", the words "Cop" and "COPenhagen", and the words "Line" and "GasoLINE". In some cases I'm sure that the software might recognize the command just due to the word containing the same pronunciation as the smaller word on its own. However, in many cases, like two of the previous examples, or with the words "Trophy"(Trow-Fee) and "Atrophy"(Ah-Truh-Fee), they are not exactly pronounced the same way in both instances. Is there a way for the program to still recognize those instances as saying the command word?

[smirgol]

I did not code most of the tool, but to my understanding what it does is to determine which word(s) you said and compare that against a list of keywords that are used in all commands, where it is looking for an exact match.

One can increase the threshold on commands to make the comparison more "fuzzy", where in general the longer the command, the higher the threshold probably needs to be, where a command should not be too long (2-3 words maybe).
If threshold is too high, it will trigger the command too easily, even by a click on the mouse or keyboard. :)

So, I'd say yes and no to your question. It is not designed to do what you ask for, but it might be achieved by increasing the threshold on the command. Then there is a Pull-Request pending where you can add multiple keywords/phrases to a command, which might be what you need.