Add multimodal RAG use-case

[jruokola]

Add support for multimodality in the agent RAG example and a guide how to extend the solution to a multi-agent setting for better results

[mukitmomin]

Thank you for the suggestion to add support for multimodal RAG use cases and extend the solution to a multi-agent setting. We appreciate you taking the time to provide this feedback, though the original issue description was a bit vague, so I want to make sure I understand the specific requirements you have in mind.

Multimodal Capabilities Regarding the multimodal capabilities, I want to let you know that support for multimodal inputs like images and text is planned for a future release of the Generative AI Application Builder on AWS (GAAB). While the current version only supports text-to-text use cases for RAG, non-RAG, and agent-based chat applications, we are actively working on enhancing the solution to include multimodal functionality.

I don't have an exact timeline for when the multimodal capabilities will be available, but I can assure you that it is a development priority for the team. If you have specific multimodal use cases in mind, I would encourage you to provide those details, as it will help us better understand the requirements and ensure we deliver a solution that meets your needs.

Multi-Agent Support In terms of the multi-agent setting, the GAAB does provide the ability to leverage multiple agents by integrating with the Bedrock Agent Collaboration feature. Customers can configure the multi-agent setup on the Bedrock console and then follow the GAAB documentation to deploy an agent use case. You can follow the guide here for more details on multi-agent collaboration using cost-effective RAG.

While we don't currently have a dedicated workflow within the GAAB for building multi-agent use cases, this is something we are evaluating for potential future enhancements. As the GAAB continues to evolve, we will assess adding more advanced multi-agent capabilities based on customer feedback and requirements.

Please feel free to provide any additional details or clarification on the specific multimodal or multi-agent needs you have in mind. We appreciate you taking the time to share this suggestion, and we will consider it as we plan future improvements to the GAAB solution.