Learn how to build a Signals-powered AI agent with AWS Bedrock AgentCore

Customer-facing AI agents are most effective when they understand what a user is doing right now - not just what they type into a chat window. Traditional agents respond generically, forcing users to repeat preferences and explain context that their browsing behavior has already revealed.

This accelerator shows you how to combine Snowplow Signals with AWS Bedrock AgentCore to build an AI agent that personalizes responses based on real-time behavioral data and persistent memory. The agent uses the Strands Agents framework and runs in a Jupyter notebook environment.

The code examples use a travel domain, but the pattern applies to any customer-facing agent - support, shopping, advisory, or content recommendation.

By the end of this accelerator you will have:

• An AI agent built with Strands Agents and AWS Bedrock
• Behavioral attributes defined and published via Snowplow Signals
• Persistent customer memory using AgentCore Memory
• An agent that combines behavioral context and memory to deliver personalized responses

The accelerator takes approximately 1 hour to complete. All source code is available in the accompanying notebook.

Architecture

The accelerator combines three components:

• Snowplow Signals processes raw event data into behavioral attributes served via the Profiles API. The agent fetches these at runtime to understand what the user is doing right now.
• AgentCore Memory provides managed short-term and long-term memory, automatically extracting preferences and facts from conversations so the agent can build context over time.
• Strands Agents is an open-source Python framework for building AI agents with custom tools and foundation models.

Together, Signals provides real-time behavioral context while AgentCore Memory provides historical context - enabling the agent to personalize responses based on both what the user is doing now and what is known from past interactions.

Prerequisites

• Signals Sandbox account or a Snowplow CDI pipeline with Signals enabled - the agent fetches behavioral attributes from the Signals Profiles API
• An AWS account with Amazon Bedrock access and AgentCore Memory access - the agent runs on Claude via Bedrock, and your IAM user needs permissions for bedrock:InvokeModel, bedrock-agentcore:*, and iam:PassRole (scoped to bedrock-agentcore.amazonaws.com) to create AgentCore Memory resources
• AWS CLI installed and configured - used to authenticate with AWS services from the notebook
• Python 3.11 or later - required by the Strands Agents framework
• Familiarity with Python and running Jupyter notebooks