Agentic AI Reference Architecture

Observability & Evaluation · Interface

Metrics Endpoint

InterfaceObservability & EvaluationObservability & EvaluationVariation point (abstract)arc:MetricsEndpoint

An HTTP endpoint (conventionally /metrics) on an instrumented agent or service that exposes its counters, gauges, histograms and summaries in a scrapeable text format.

Responsibility. Exposes a component's current metrics for collection.

Also known as: /metrics endpoint, Metrics exporter endpoint

is exposed byis invoked byis exposed byis specialized byAgent Controller: is exposed byAgent ControllerMetrics Collector: is invoked byMetrics CollectorGPU Telemetry Exporter: is exposed byGPU Telemetry ExporterInference Metrics Endpoint: is specialized byInference Metrics Endpoint
Direct neighbourhood (hover for relationship types)

Variants

VariantWhen to choose
Inference Metrics Endpoint—

Relationships

is exposed by structural

is invoked by dependency

Design guidance

Classification

Patterns
Pull-based metrics exposition
Technologies
Prometheus text formatprometheus_client
Quality attributes
Maintainability (ISO/IEC 25010)

Sources

  1. Ch4.1: T. Nguyen, "Introduction to AI Agent Deployment and Scaling," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 4.1. ISBN: 9798244538229.
  2. Ch8.1: T. Nguyen, "Latency Metrics," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 8.1. ISBN: 9798244538229.