Deployment guide

Deploy the Customer 360 solution in 5 phases using the provided CDK stacks.

Prerequisites

AWS account requirements: * Admin access or equivalent permissions * Quick Suite Enterprise subscription * Bedrock model access (Claude 3.5 Sonnet) * Service quotas for S3, Glue, Athena, Aurora

Local environment: * AWS CLI configured * Node.js 18+ installed * Python 3.9+ installed * CDK CLI: npm install -g aws-cdk

Deployment phases

Phase 1: Data Lake Foundation (10 minutes)

cd guidance-for-agentic-customer-360
make phase1

Creates S3 bucket, Glue database, Athena workgroup, Lake Formation permissions.

Phase 2: ETL & Prerequisites (5 minutes)

make phase2

Creates Glue crawler, IAM roles, Quick Suite data source.

Phase 3: Data Generation (20 minutes)

make phase3

Generates 500K customers, 1.4M interactions, 8 Athena views.

Phase 4: Quick Suite Dashboards (15 minutes)

make phase4

Creates 8 datasets, dashboard, demo user.

Phase 5: Bedrock AI Agent (30 minutes)

make phase5

Deploys Aurora cluster, Knowledge Base, Bedrock Agent, Lambda functions.

Verification

Check data:

aws s3 ls s3://automotive-cx-data-lake-<ACCOUNT-ID>/raw/
aws athena start-query-execution \
  --query-string "SELECT COUNT(*) FROM customers" \
  --query-execution-context Database=cx_analytics

Check Quick Suite:

aws quicksight list-data-sets --aws-account-id <ACCOUNT-ID>
aws quicksight list-dashboards --aws-account-id <ACCOUNT-ID>

Check Bedrock:

aws bedrock-agent list-knowledge-bases
aws bedrock-agent list-agents