You can access individual letters with the mid() expression. The tokenat() expression can be useful to grab individual words if there are spaces in between. To filter around punctuation you can remove them first with replace() or maybe regexreplace to remove all punctuation in one shot, but that can be tedious to do.