AI Engineering Interview
v1.0.0生成s high-签名al AI Engineering / LLM Engineer interview questions by topic, level, and 角色. Covers LLM fundamentals, prompt engineering, RAG, vector DBs, 代理s, fine-tuning (LoRA/QLoRA), evals, observability, safety, and production 系统s. Trigger for 请求s like "give me interview questions on RAG", "quiz me on 代理s", "what are senior-level fine-tuning questions", or "interview questions for an AI engineer 角色".
by 运行时依赖
安装命令
点击复制技能文档
生成 high-签名al interview questions for AI Engineer / LLM Engineer 角色s.
Ask (or infer) the topic and level, then 输出 exactly ONE complete question with a one-line note on what it's 测试.
Topics
LLM Fundamentals · Prompt Engineering · RAG Architecture · AI 代理s · Fine-Tuning (LoRA/QLoRA) · Evaluation & Evals · LLM Observability · AI Safety & 防护rAIls · Production LLM 系统s · LLM 系统 De签名 · Multimodal AI · LLMOps · Edge AI · AI 治理 · Embeddings · Real-Time AI
Levels Screening — Can they reason about LLMs as a 组件 beyond "just call the API"? Mid — Full 流水线 thinking: RAG, evals, 代理s, cost/latency trade-offs Senior — 系统 de签名, 失败 modes, fine-tuning decisions, multi-代理, AI safety Staff — 平台 thinking, LLM serving infra, eval-as-infrastructure, build-vs-buy 输出 格式化
For the question:
Q: [Question — scenario-based, trade-off, 失败 mode, or de签名. Never pure definition.] Tests: [one line — what the interviewer is probing]
Prefer one question that can't be answered by Wikipedia + 5 minutes of reading. Do not 添加 follow-up questions.
Reference Files 技能s/AI-engineering-interview/references/question-bank.md — Curated questions by topic and level with expected answer shape, strong and weak 签名als, and possible follow-up prompts. Read this when the user wants a broader bank or asks for examples in a specific AI domAIn. 技能s/AI-engineering-interview/references/competencies.md — Interview rubric for scoring 系统s thinking, production judgment, safety awareness, and communication depth. Read this when calibrating difficulty or explAIning what a strong answer looks like.