InterviewsVector
Capability spine

Arc 09 · Staff AI Engineer

Evaluation, Safety, and Reliability

Make quality a release system: task-specific datasets, calibrated judges, human review, security tests, telemetry, and incident learning.

Exit capability: Define a release bar that catches quality, safety, and reliability regressions.

Mapped lessons
7
Published now
0
Full-arc estimate
17 hours
Last edited
2026-08-11

Lesson sequence

Live units open into complete labs. Planned units stay visible to show the dependency path, but intentionally have no detail route.

  1. 01

    An Eval Is a Decision System

    Planned

    Start from a release decision, then choose cases, graders, uncertainty, and escalation.

    Design reviewAdvanced105 min estimate

  2. 02

    Build a Golden Dataset That Can Disagree with You

    Planned

    Sample real tasks, hard negatives, long tails, and policy boundaries without teaching to the test.

    Build labAdvanced110 min estimate

  3. 03

    Calibrate LLM Judges

    Planned

    Measure position bias, verbosity bias, self-preference, variance, and disagreement against human labels.

    Failure labAdvanced115 min estimate

  4. 04

    Online Evaluation Without Shipping Blind

    Planned

    Connect shadowing, canaries, holdbacks, user signals, and rollback thresholds.

    Systems labAdvanced105 min estimate

  5. 05

    Observe the Decision Path, Not Just the Model Call

    Planned

    Trace retrieval, prompts, tools, model versions, policy checks, latency, and cost with controlled data capture.

    Systems labAdvanced110 min estimate

  6. 06

    Build a Safety Case for an AI Feature

    Planned

    Tie hazards to controls, evidence, residual risk, ownership, and launch conditions.

    Design reviewAdvanced105 min estimate

  7. 07

    AI Incident Response and Correction Loops

    Planned

    Contain harm, preserve evidence, distinguish model and system causes, and turn findings into durable gates.

    Failure labAdvanced110 min estimateArtifact: AI incident playbook