When looking at the history of recordings and the AI output. Create option to launch a chat window with the transcription and the AI output. This would allow for deeper investigation right in the application, vs having to take the output into another model to do it. This is useful for questions that are outside the scope of the original prompt, which is normally in my case used to pull out speakers, summarize and pull out key takeaways. Sometimes I want to do ad hoc prompts on the data which could yield more insights.