EcommercePlatformMCPAgentCore

Set up Knowledge Base in the console. knowledge base will later be accessed by the MCP Server (Module 3) and ultimately by the Strands Agent (Module 4) to provide context-aware product assistance in the chatbot. 1-Open Amazon Bedrock.

1-Give your knowledge base a name (or keep the default). 2-For IAM permissions, select Create and use a new service role 3-Select Amazon S3 as the data source. 4-Select Next.

1-Under Advanced settings, keep default KMS key. 2-Select Next.

After creation, ensure the Knowledge Base is using the latest version of the S3 source:

1-In the Knowledge Base details page, select the S3 Data Source. 2-Select Sync. 3-Copy and save the Knowledge Base ID locally. We will use it in Module 3.

Setting up the MCP Server Run Jupyter notebook for creating MCP server 1-In the AWS Console, open Amazon SageMaker AI. 2-In the left navigation, select Applications and IDEs → Sagemaker Studio. 3-Under User Profiles, select default-user, then select Open Studio.

1-In SageMaker Studio, open the Applications pane. 2-Select JupyterLab. 3-For ecommerce-workshop-space, select Run. When the status is Running, select Open.

1-Switch to Module 3 in your Jupyter Notebook. 2-Open hosting_mcp_server.ipynb 3-Ensure the kernel selected is Python 3 (ipykernel). 4-Run each cell step-by-step

1-Switch to Module 4 in your Jupyter Notebook. 2-Open hosting_agent.ipynb 3-Ensure the kernel selected is Python 3 (ipykernel). 4-Run each cell step-by-step.

go to the wesite (https://d24fs6fbvcdlda.cloudfront.net/) From the left navigation menu, click GenAI Chatbot.

How many running shoes do you have in stock? Place an order for one pair of running shoes for me.