Technical interview guide

Prompt Engineering & Prompt Management

Designing reliable prompts and treating them as versioned, tested production artifacts rather than one-off strings.

Read: 45 min
Practice MCQs: 25
Interview QA: 25
Edition: v4
Editorial status: Reviewed
Relevant for: AI EngineerAgentic AI EngineerForward Deployed Engineer (FDE)

Scope: GPT-3 few-shot prompting, chain-of-thought, self-consistency, ReAct, PAL, PromptSource, DSPy, indirect prompt injection, HELM, Model Cards, JSON Schema 2020-12, NIST GenAI, and OWASP Prompt Injection references reviewed 2026-09-06.

Interview QA

Treat each question like a live interview question: answer out loud first (structure, assumptions, tradeoffs), then open the model answer to spot gaps and rehearse a tighter follow-up.

Curated: 2026-10-05 · Written: 2026-10-05 · Reviewed: 2026-10-03

QA-1

Write a behavior contract for a prompt-backed feature.

QA-2

Design a production prompt-template schema.

QA-3

Test a prompt renderer at its input boundary.

QA-4

Design instruction hierarchy and conflict handling.

QA-5

Select and govern few-shot examples.

QA-6

Evaluate stability of a few-shot prompt.

QA-7

Prevent template injection during prompt rendering.

QA-8

Use reasoning prompts for a multi-step task safely.

QA-9

Decide whether to use self-consistency.

QA-10

Build a reliable extraction prompt and validator.

QA-11

Design a structured-output repair loop.

QA-12

Threat-model direct prompt injection.

QA-13

Defend a RAG assistant against indirect injection.

QA-14

Design secret handling for an LLM application.

QA-15

Apply least privilege to prompt-driven tools.

QA-16

Design context budgeting for a prompt template.

QA-17

Design prompt versioning and promotion states.

QA-18

Build a prompt-change evaluation pipeline.

QA-19

Roll out a prompt version safely.

QA-20

Design safe caching for prompt-backed responses.

QA-21

Instrument prompt behavior in production.

QA-22

How do you detect, isolate, and mitigate sudden model output degradation or hallucination spikes in production prompts?

QA-23

Migrate a prompt stack to a new model.

QA-24

Operate prompt retirement and deletion.

QA-25

What key architectural components are required in an enterprise prompt registry to support prompt versioning, RBAC, CI/CD eval gates, and environment promotion?