How to query all of the data/files in my S3 bucket

Question

Hi all,

I found the following repository on GitHub https://github.com/aws-samples/rag-using-langchain-amazon-bedrock-and-opensearch.git that codes for a genAI AWS Bedrock chatbot using your own data.

I am trying to replicate the chatbot using the dataset provided in the repository, just so I can confirm I've got everything working well on my end before inputting my own data.

I've noticed that the data is in the form of a URL in this code. How do I adjust the code such that it retrieves data from my S3 bucket, rather than a URL? Also not 100% sure which lines of code in the .py files in the repository are retrieving data from the URL.

Thank you very much (:

Didier_Durand · Answer

Hi,

You now have a much simpler way to achieve this: It's Knowledge Bases for Amazon Bedrock: https://aws.amazon.com/bedrock/knowledge-bases/

It will manage all the RAG mechanism for you as it is a managed service: you just have to put the data in a bucket and Bedrock KB will do the rest for you.

If you want to better understand how this KB RAG works, it's fully detailled in my previous article: https://repost.aws/articles/AR-LV1HoR_S0m-qy89wXwHmw/the-leverage-of-llm-system-prompt-by-knowledge-bases-for-bedrock-in-rag-workflows

Best,

Didier

How to query all of the data/files in my S3 bucket

Relevant content