Workshop Guide

Horizon Media · Observability velocity · Live OTel on Elastic Serverless. Use the Elastic Serverless tab + your AI assistant.

Scorecard template (CSV)

Challenge 1: Explore live metrics in Elastic Serverless, then run prompts 1 → 2 in your AI assistant. Pick your #1 at-risk or stalled service for Challenge 2.

Serverless — explore first

Elastic Serverless tab → **Metrics**
Stream: metrics-generic.otel-default · filter service.name:horizon-*
Skim 3–5 series (e.g. http.server.requests, errors, heartbeats)

Velocity signals

Accelerating — ingest trending up, healthy request rates, dashboards usable
At risk — error spikes, cardinality noise, or missing SLO coverage
Stalled — flat ingest, no alert/runbook progress, POV stuck

1. Capture Serverless snapshot

Run first — paste what you saw

Horizon Media — Observability velocity. You are my observability adoption coach.

Scenario: **Horizon Media Streaming** — a customer evaluating Elastic Observability on Serverless.
I explored Metrics Explorer in my lab project (OTel on metrics-generic.otel-default).

I will paste 3–5 observations from Serverless (service names, metric names, trends I saw).
Do not invent metrics — wait for my paste.

After I paste, confirm: "✓ Serverless snapshot loaded — ready for velocity scoring."

2. Observability velocity scorecard

Challenge 1 main exercise

Using ONLY my Serverless observations from the previous message — map velocity signals across Horizon services.

Output a markdown table:

| Service | Key metric | Trend / signal | Velocity signal | Why |
|---------|------------|----------------|-----------------|-----|

Velocity signal = **Accelerating**, **On track**, **At risk**, or **Stalled**

Rules:
- **Stalled** if ingest flat or no meaningful series for the service
- **At risk** if errors elevated or noisy/high-cardinality patterns dominate
- **Accelerating** if request rates healthy and adoption milestones visible

Sort by risk (Stalled first). Highlight top 3 at-risk services.
Pick **#1 service** for Challenge 2 — best candidate for a time-to-value narrative.
Cite metric names and trends from my Serverless paste only.

3. #1 service deep dive

Optional — Challenge 2 has full narrative

For my #1 at-risk or stalled service from the scorecard:

Using Serverless facts from our chat only:
1. What the metrics pattern implies for the POV
2. Likely blockers (noise, missing dashboards, no alerts)
3. One sentence: what changed since kickoff?

Under 150 words — or continue in Challenge 2 lab guide.