gen-ai · k8s
Kube + LLM, made readable
Learn
All paths
Practical · observability

Observability for LLM systems

You will know which metrics actually matter (TTFT, TPOT, goodput) and how to wire logs, metrics, and traces for streaming workloads.

Progress
0 / 3 lessons
Start
  1. 01
    Metrics that matter for LLM serving
    TTFT, TPOT, throughput — and why CPU/memory miss the point
  2. 02
    Building the observability pipeline
    Logs, metrics, traces — for a workload that streams
  3. 03
    Quality, guardrails, and hallucination detection
    When 'green dashboards, wrong answers' becomes the failure mode
Lock it in
Detecting a quality regression

You are about to roll out a new fine-tune. Infra metrics are green.

Try the scenario