Skip to main content

Metadata

Metadata in VoltOps tracing consists of key-value pairs that provide rich context and organization for your AI workflows. Unlike the core functional data (inputs, outputs, errors), metadata helps you organize and understand the business context, operational environment, and relationships within your AI system.

By adding structured metadata to your traces, you can categorize, filter, and analyze your AI applications with precision, making debugging and optimization significantly more effective.

Metadata Usage Patterns

Basic Agent ID
User & Session Context
Organizational Context
Multi-Agent Relationships
LLM Model Parameters

Basic Agent ID

Simple agent identification for better tracking and organization.

JS/TS SDK
Vercel AI SDK
Python SDK
Example:
const trace = await sdk.trace({
  name: "Customer Support Query",
  agentId: "support-agent-v1",
  input: { query: "How to reset password?" },
  metadata: {
    priority: "high",
    source: "web-chat"
  }
});

Technical metadata helps optimize model performance, track different configurations, and correlate AI behavior with specific parameter settings. Essential for prompt engineering and model tuning.

Why Metadata Important​

Metadata enables sophisticated filtering across your traces:

  • Find all traces from a specific user or conversation
  • Filter by agent type, department, or priority level
  • Locate workflows in specific environments or projects
  • Search by custom business categories

Performance Analysis​

Organize performance data by meaningful business dimensions:

  • Compare agent performance across different user segments
  • Analyze costs by department or project
  • Track improvements across agent versions
  • Monitor environment-specific metrics

Debugging and Troubleshooting​

Metadata provides essential context for debugging:

  • Quickly identify the business context of failed traces
  • Correlate issues with specific user patterns or environments
  • Track problems across related agents in complex workflows
  • Understand the impact of issues on different user segments

Compliance and Auditing​

Structured metadata supports governance requirements:

  • Maintain audit trails with business context
  • Track data processing by user consent levels
  • Associate AI decisions with responsible teams
  • Monitor compliance across different regulatory environments

Table of Contents