Enable data capture on the inference endpoint by adding a DataCaptureConfig to the endpoint configuration, specifying a destination S3 URI and a capture percentage
Create a baseline from a representative dataset using a DefaultModelMonitor.suggest_baseline() job, which generates constraint and statistics JSON files in S3
After scheduled executions run, retrieve results with monitor.list_executions() and inspect violation reports in the output S3 path
Set up a CloudWatch alarm on the metric aws/sagemaker/Endpoints/data-metrics to receive notifications when constraint violations exceed a threshold
Stop the monitoring schedule when the endpoint is decommissioned: monitor.stop_monitoring_schedule()
Known gotchas
Data capture must be enabled before the monitoring schedule is created and must capture at least a few hundred requests before the baseline comparison is statistically meaningful
Model Monitor uses a built-in container that runs on ml.m5.xlarge by default; the instance type and schedule frequency drive ongoing cost even when no violations occur
The baseline suggest_baseline() job and the monitoring execution jobs run as separate processing jobs — IAM role must have s3:GetObject and s3:PutObject on both the capture and output buckets
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp