RAG on Bedrock

Setup to run your own instance

1. Create Streamlit Secrets The following attributes must be provided in your secrets.toml. See Streamlit docs fro more help

   • ASTRA_VECTOR_ENDPOINT
   • AWS_ACCESS_KEY_ID
   • AWS_SECRET_ACCESS_KEY
   • AWS_DEFAULT_REGION

   If you want to use LangSmith, then these are also required:

   • LANGCHAIN_ENDPOINT
   • LANGCHAIN_API_KEY

2. Run the app

   streamlit run app_bedrock.py

Using the demo

1. Provide content

   • Use the sidebar to provide content that will be referenced in the RAG workflow.
   • You can choose from:
     • Uploading a file
     • Providing a URL
     • Providing text

2. Select your preferred LLM from the available Bedrock Foundation Models

3. Ask a question