We have been exploring the use of GPT-4 Vision for Multimodal RAG, and I’ve written up a blog post on an insurance industry use case.
https://www.graphlit.com/blog/multimodal-rag-insurance-insights
Looking forward to the production models so we can release our integration.