AI Observability Module Installation

The AI Observability module is delivered as three complementary components.

---

Delivery Composition

Component	Purpose
Content module	Dashboards, scheduler tasks, runtime indexes, tags, ISM policies
Utility	Deployment of index templates, index patterns, GPU inventory asset, and health metrics
Data collection configurations	Data collection configurations for LLM-Gateway, Inference, Claude Code, Codex, GPU

---

Installation Steps

1. Installing the content module — importing the ai_observability.zip content module (dashboards, tasks, indexes, ISM policies, tags)
2. Installing the utility — ai_observability_ko_maker utility for deploying objects not included in the content module
3. Data collection setup — deploying data collection configurations and setting up telemetry collection routes

Installing the Content Module

Importing the AI Observability content module through the Smart Monitor Content Management menu.

Utility Installation

Deploying infrastructure objects of the AI Observability module using the ai_observability_ko_maker utility.

Data Collection Setup

Deploying the AI Observability module Data Collection Configuration — SAF Beat Manager, Logstash, OTLP Collector, LiteLLM gateway.