Instructor Delivery Guide
This page is the instructor hub. Each track has its own one-page datasheet and its own delivery guide, delivered as PDFs:
| Track | Datasheet | Delivery guide |
|---|---|---|
| AgentOps Briefing (~1 hour) | Briefing datasheet (PDF) | Briefing delivery guide (PDF) |
| AgentOps VBD Workshop (~8 hours) | VBD datasheet (PDF) | VBD delivery guide (PDF) |
The PDFs are the per-track source of truth for run-of-show, timing, and facilitator cues. The notes below are shared guidance that applies to both tracks.
Delivery intent
The workshop should feel like a production-readiness engagement, not a product tour. The instructor should repeatedly connect concepts back to the AgentOps four-pillar operating model:
Evaluate -> Ship -> Observe -> Operate
Recommended delivery style
- Start from the production question: “Can we safely ship this agent?”
- Use one production-candidate agent as the thread through the session.
- Treat observability as an engineering discipline, not as screenshots of dashboards.
- Use failures intentionally. A failed eval or release gate is the clearest way to show value.
- Keep AgentOps Toolkit references subtle and practical.
For AgentOps Briefing
AgentOps Briefing should prioritize:
- Production gap and AgentOps operating model
- Evaluation and release gates
- Observability for agents
- Demo video
- 30-day starting path
The observability section should explain trace correlation and the trace-to-evaluation feedback loop.
For AgentOps Value Delivery Workshop
AgentOps Value Delivery Workshop should prioritize:
- One consistent scenario
- One readiness artifact per lab
- Strong observability thread across labs
- Capstone release-readiness review
The dedicated observability lab should be treated as a major part of the day, not a side topic.
Instructor checklist
Before delivery:
- Confirm the target audience and expected technical depth.
- Confirm whether the session is AgentOps Briefing or AgentOps Value Delivery Workshop.
- Confirm whether live demos, recorded demos, or screenshots will be used.
- Confirm the Foundry project and agent scenario.
- Confirm that no sensitive telemetry or tenant data appears in screenshots.
- Confirm the GitHub Pages site builds after content changes.
After delivery:
- Capture questions that should become FAQ entries.
- Capture observability gaps that should become lab improvements.
- Capture production scenarios that should become evaluation examples.