AWS · MLA-C02 · BETA
Observability, dashboards, and agents
MLA-C02 · Lesson 20 of 23 · 2:44:22
YouTube loads only after you press play. This chapter opens at its timestamp in the full course. Watch on YouTube ↗
In this lesson
Lesson 20 of the free MLA-C02 beta series (AWS Certified Machine Learning Engineer – Associate, domain 4, operating, monitoring, and securing). In this lesson: AgentCore Observability, X-Ray, and CloudWatch Logs Insights, tool failures, truncated answers, coordination problems, and loops, CloudWatch dashboards and Amazon Quick, and the consumption of each agent. With three quick checks and a recap, and AnyCompany Mobility, a car sharing service, as the running example.
Preview the topics
- Some answers from the support agent stop in the middle of a sentence. The traces show the stop reason max_tokens on those calls. What should the ML engineer do first?
- Regional managers want a weekly view of trips, revenue, and the accuracy of the demand forecast, with filters by city. They never use the AWS console. What should the ML engineer build?
- After a prompt change, the agent's tokens per session doubled, but the number of resolved chats stayed the same. Which metric should the team watch to catch this kind of change early?