← Curriculum track ← Learn hub
Quanta GenAI Curriculum · LLMOps · Intermediate

LLMOps Intermediate — 051: stress-test tool permissions on `Eval harness hardening` — memo `862897 [51]`

Lesson 051: Eval harness hardening

Focus

Bias toward observable metrics: latency, cost, escalation rate. Token Eval harness hardening:51 keeps neighbouring lessons differentiable.

Key ideas

Deep dive notebook

Synthetic drill artefacts

Token CFO scratchpad

- prompt_budget: 2014
- completion_budget: 500
- cache_key: `64be`

Hypothesis: halving completions moves P95 ~8% — record actuals.

Practice

Practice Draft three eval assertions QA must pass before prompt promotion. — 51 Bump 3.