Finds the real topics
It reads every fragment, not just the keywords, and groups thoughts by what they are actually about. A note about a 1am usage spike and a note about losing people after a month end up together, because they are both about retention.