I would suggest first defining what you think could use improvement. Training objectives, key learning, layout, subject organization, writing style for clarity or to add enthusiasm/interest, student interaction, etc. Training should consider verbal, visual, and kinesthetic learning styles for different types of users. Some course developers focus on words while others keeping their students engaged and each can often use improvements in effective teaching.
Different types of AI such as language vs. graphics focused can be addressed with different tools. In fact if you were to type your question into a generative tool like Chat GPT, it will probably give you around 8 suggestions for different use cases. Ask specific questions about each use case and you can get another layer of detail. Whether or not that information is actually accurate or useful is another question altogether. Saving Changes...
If you provide the Gen AI tool with an understanding of who the audience is, what your objectives are, what tone you wish to deliver the presentation in and other constraining factors in your prompt and ask it to ask you additional questions you are likely to get a decent response.
You might also look at combining AI tools - ChatGPT for the text content and one of the multiple visual presentation-focused tools for that aspect.