Hi @reignM,
Your regex pattern is on the right track! There are a variety of patterns you could use. If you use the regex string functions in the formula tool, here are two of my favorite ways to solve this use case.
This first regex pattern is instructing Alteryx to replace any instance of a non-uppercase character (can use \U or [^A-Z] to represent a non uppercase character) with a blank string. In effect, this will keep only the uppercase characters. For "Video of JKL", it would keep "VJKL", so it depends on how you want it to behave. The last argument, 0, instructs the pattern to be case sensitive. By default, most of the regex functions in Alteryx are case INsensitive.
regex_replace([StringColumn], '\U','',0)
The second pattern says to capture and return the first group of uppercase characters that have 2 or more occurrences. This pattern can break more easily or may not give the right results if you have multiple matching groups in your string.
regex_replace([StringColumn], '.*(\u{2,}.*','$1',0)
Like I mentioned before, there are a variety of ways to get a functional pattern in regex. I'd encourage you to experiment until you find one that works for you! Rubular is my favorite site for testing regex patterns. Heads up that regex functions are computationally expensive, so if you find another way to extract your desired strings, that will be easier on your computer with a large data set.
I hope this helps!
