Define the lab goal and success criteria

This lab builds OpenTelemetry traces across retrieve → rerank → generate for support answers. The human in the loop is on-call engineer debugging a slow groundedness miss. Scope is intentionally narrower than “make AI reliable”: you will prove one oracle — fixture question produces 3 child spans; total latency attributes sum within 5% of root — and one invariant — every span carries trace_id + redacted attrs; prompts sampled ≤ 1%. Record non-goals in your notes so a later change cannot silently expand authority. The incident mnemonic for the chapter is TRACE-ORPHAN-441; design as if that ticket is already written and you are filling evidence.

Turn the oracle into a table: input fixture, expected observable, prohibited side effect, owner, latency/cost ceiling. Separate model taste from software correctness — transport, auth, parsing, and termination must be deterministic even when generated text varies. Primary metric family: trace_completeness ≥ 0.99 and p95_end_to_end_ms ≤ 2500. Averages without a denominator or revision label do not gate release. Fake external dependencies in unit tests; live calls wait until fakes pass.

Before implementation, encode a failing check that would have caught missing parent context → orphan spans, cannot join retrieve to answer. That failure is the pedagogical north star for later pages: contracts reject it, happy path never performs it, validation asserts it, failure-handling contains it, observability detects it, security-ops prevents privilege tricks around it, and mastery replays it in a drill. Endpoint under study: POST /v1/support/answer.

Capacity note for planners: estimate peak demand on POST /v1/support/answer and the cost ceiling for a failed retry storm. Write the abort conditions — unbounded spend, cross-tenant leakage, or inability to roll back — before you enjoy the first green test. Prefer synthetic fixtures shaped like production over anonymized production dumps you cannot share in class. When you are tempted to widen scope, re-read the oracle (fixture question produces 3 child spans; total latency attributes sum within 5% of root) and cut features that do not serve it. The teaching outcome is judgment under constraints: on-call engineer debugging a slow groundedness miss gets a trustworthy control, not a kitchen-sink framework. Keep the language of release decisions: promote, hold, or roll back — never “see if it gets better.”

Decide what will live in version control on day one: fixtures, contract markdown, and a failing test name. Write the cost ceiling as a hard number with currency and period. If the lab involves clusters, name the non-prod context you will use and forbid prod kubecontexts in scripts. Capture the baseline metric once before changing code so later gains are comparative. Refuse tools that hide the request path behind magic macros until the oracle is green on fakes. Your README section for this page should be five lines or fewer and still falsifiable. In this chapter the product is OpenTelemetry traces across retrieve → rerank → generate for support answers, the human stakeholder is on-call engineer debugging a slow groundedness miss, and the incident id you design against is TRACE-ORPHAN-441. Re-state the oracle in your notes — fixture question produces 3 child spans; total latency attributes sum within 5% of root — and keep the invariant visible: every span carries trace_id + redacted attrs; prompts sampled ≤ 1%. Track trace_completeness ≥ 0.99 and p95_end_to_end_ms ≤ 2500 as the scoreboard. Surface under change control: POST /v1/support/answer.