Uploaded July 2026 | Updated September 2026, 2 weeks ago

---

Learn how to set up alerting that catches agent failures fast, without paging you for every minor blip. Evals produce scores, but a monitor is what tells you something's actually wrong without you watching a dashboard.

The difference between a useful alert and noise is routing: guardrail failures should page someone now, while everything else can wait for a weekly review.

Watch this to learn:
• How to configure a monitor to watch an eval label, with static vs. automatic thresholds
• Where alerts should go (Slack, PagerDuty) and how to route guardrails vs. weekly reviews
• Why evals also belong in your CI/CD pipeline, running on every prompt and code change

Chapters:
00:00 Why monitors matter
00:15 What a monitor watches
00:52 Configuring a monitor and thresholds
02:08 Where alerts go, Slack and PagerDuty
02:33 Route guardrails vs. weekly reviews
02:58 The full production loop
03:49 Evals in your CI/CD pipeline
04:27 Next: closing the loop

👉 Sign up for free: app.arize.com/auth/login?utm_source=youtube&utm_medium=organic_social&utm_campaign=arize_ax_getting_started
🔗 Learn more about Arize AX: arize.com
📓 Colab notebook for the series: colab.research.google.com/drive/1dViThD0kJjbqtDGE7-ciBYIRwV-HBN1T
📚 Docs: docs.arize.com
🔔 Subscribe for weekly content on LLMs, agents, and evaluation: youtube.com/@arizeai?sub_confirmation=1

#ArizeAX #AIMonitoring #AgentObservability