Text2Slide multiclass classification

I am considering an idea of stitching together a slide deck based on text input, e.g. given: An all-hands presentation with business updates, project timelines, and financial report charts the output could be a deck with slides corresponding to Title, List, Calendar, Pie Chart, Conclusion.

I have preexisting slides that are mostly categorized by the form ranging from very general like List to more specific like Decision Tree or Venn Diagram.

Am I on the right track that this sounds like a multiclass classification problem? Maybe it could be multi-label, as something like a Customer onboarding text intent could be matched to both a linear Funnel slide or a Decision tree slide.

Apologies if this question is misguided, I'm a dev out of their element. Very new to this, should I run with python+jupyter+scikit-learn and just #figureitout?