I’m currently working on this with the caveat that I want to do the work locally. Using whisper but the summarization portion if this task is not straightforward given the limited context size of models.

Does anyone have any additional insight into this problem?