Orchestration · Software component

Event-Triggered Agent

Software componentOrchestrationOrchestration & Toolsarc:EventTriggeredAgent

A stateless agent function that stays dormant until a subscribed event arrives, reconstructs context from the event payload and external stores, acts, and emits result events for downstream agents.

Responsibility. Performs one specialised agent step in response to a subscribed event type.

Also known as: Serverless agent, Agent function, Event-driven agent

Direct neighbourhood (hover for relationship types)

Relationships

reads dependency

writes dependency

emits telemetry to dynamic

Trace Collector Ch4.2

is routed to by dynamic

Content-Based Event Router Ch4.2
Rollout Manager abstract Ch4.2

is triggered by dynamic

receives data from dynamic

sends data to dynamic

Event Broker abstract Ch4.2

is constrained by control

is orchestrated by control

Durable State Machine Orchestrator Ch4.2

is monitored by assurance

Stalled Workflow Resumer Ch4.2

Design guidance

MUST obtain all required context from the triggering event or external storage, never from in-memory state persisted between invocations.
MUST process events idempotently by recording processed event IDs, because delivery is at-least-once.
SHOULD place expensive initialisation (model loading, database connections, configuration) outside the per-request handler so warm invocations reuse it.
SHOULD be invoked asynchronously rather than synchronously from the API gateway, and SHOULD NOT turn naturally synchronous queries into events.
SHOULD save progress to durable storage and exit gracefully before the platform timeout for long-running workflows.

Quantitative guidance

As stated by the sources; verify before use.

Lazy-loading a 2 GB model keeps cold starts under 1 s versus 30+ s when loaded on every cold start (Ch4.2).

Classification

Patterns: Event choreography (no central orchestrator)Initialise outside the handler (global-scope caching)Lazy loadingIdempotent consumerAsynchronous invocationCheckpoint-and-continue
Technologies: AWS LambdaAzure FunctionsAmazon BedrockAmazon SageMaker endpointsAmazon SES
Quality attributes: Maintainability (ISO/IEC 25010)Performance efficiency (ISO/IEC 25010)
Risks mitigated: Context loss across stateless invocationsDuplicate side effects on event redeliveryCold-start latencyTight coupling between event producers and consumers

Sources

Ch4.2: T. Nguyen, "Deployment and Scaling," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 4.2. ISBN: 9798244538229.