CPN Learning Path CPN Connect On Demand Library All Courses Home

Administration and Measurement

Most Day 1 failures aren't technical, they're the wrong role assigned or the spend limit still at zero. This course covers seat types, role hierarchy, the Analytics and Compliance APIs, OpenTelemetry pipeline setup, audit logs, and per-team cost attribution.

rate limit

Code not recognized.

About this course

Administration is what makes the deployment actually work, and most of it happens in the two weeks before kickoff, not after. This course is built around a simple insight: the most common Day 1 failures aren't technical. They're the wrong role assigned to the IT contact, the spend limit still at zero, or a seat type that blocks Claude Code access entirely. You can catch all of them in advance.

You'll start with seat types and the role hierarchy:

  • Seat types across legacy and usage-based billing models, and which ones actually include Claude Code access (Standard seats do not)
  • The four roles; Primary Owner, Owner, Admin, and User, and specifically the gap between what 'Admin' sounds like it can do and what it actually can (Admin cannot configure SSO or provision seats)
  • The pre-kickoff checklist: four steps, in the right order, that prevent the escalations that otherwise hit on Day 1

From there, the course moves into the measurement infrastructure that makes a Day 30 readout defensible:

  • The Analytics API: pulling usage and adoption data, and knowing which metrics hold up in an executive conversation versus which ones are vanity numbers
  • The Compliance API: near-real-time prompt inspection, SIEM integration, eDiscovery and legal hold, the four use cases security and compliance teams ask about
  • OpenTelemetry pipeline setup: the reference configuration, span attributes, and how to design a three-metric dashboard against the activation scorecard
  • Audit logs: what's captured, how to query, and what a compliance team needs to see
  • Cost monitoring and attribution: how to set spend limits, build per-team chargeback using OTel resource attribute labels, and have a sensible conversation with finance about AI spend

Practice: work through the supplied OTel reference config and design a three-metric dashboard. If you have admin access to a Claude Enterprise tenant, stand up the pipeline there. Your artifact is the OTel config plus a sample dashboard spec.

Time In Skilljar (Measured) 1 Hour 49 MinutesPractice & Artifact Building (Estimated Time In Skilljar) 55 Min

Curriculum

  • Seat management and admin console
  • Analytics API for usage and adoption
  • Compliance API for audit and oversight
  • OpenTelemetry pipeline setup
  • Audit logs and querying
  • Cost monitoring, attribution, and spend controls

About this course

Administration is what makes the deployment actually work, and most of it happens in the two weeks before kickoff, not after. This course is built around a simple insight: the most common Day 1 failures aren't technical. They're the wrong role assigned to the IT contact, the spend limit still at zero, or a seat type that blocks Claude Code access entirely. You can catch all of them in advance.

You'll start with seat types and the role hierarchy:

  • Seat types across legacy and usage-based billing models, and which ones actually include Claude Code access (Standard seats do not)
  • The four roles; Primary Owner, Owner, Admin, and User, and specifically the gap between what 'Admin' sounds like it can do and what it actually can (Admin cannot configure SSO or provision seats)
  • The pre-kickoff checklist: four steps, in the right order, that prevent the escalations that otherwise hit on Day 1

From there, the course moves into the measurement infrastructure that makes a Day 30 readout defensible:

  • The Analytics API: pulling usage and adoption data, and knowing which metrics hold up in an executive conversation versus which ones are vanity numbers
  • The Compliance API: near-real-time prompt inspection, SIEM integration, eDiscovery and legal hold, the four use cases security and compliance teams ask about
  • OpenTelemetry pipeline setup: the reference configuration, span attributes, and how to design a three-metric dashboard against the activation scorecard
  • Audit logs: what's captured, how to query, and what a compliance team needs to see
  • Cost monitoring and attribution: how to set spend limits, build per-team chargeback using OTel resource attribute labels, and have a sensible conversation with finance about AI spend

Practice: work through the supplied OTel reference config and design a three-metric dashboard. If you have admin access to a Claude Enterprise tenant, stand up the pipeline there. Your artifact is the OTel config plus a sample dashboard spec.

Time In Skilljar (Measured) 1 Hour 49 MinutesPractice & Artifact Building (Estimated Time In Skilljar) 55 Min

Curriculum

  • Seat management and admin console
  • Analytics API for usage and adoption
  • Compliance API for audit and oversight
  • OpenTelemetry pipeline setup
  • Audit logs and querying
  • Cost monitoring, attribution, and spend controls