First is a RAG model that is fed chunks of podcast transcript data to help answer the user's question about personal finance for themselves and their family as a whole.
The second model I plan to implement is a model fine-tuned to the data, hopefully with better performance, to accomplish the same feat of helping the user with their personal and family finance problems.