// the find
amitshekhariitbhu/ai-engineering-interview-questions
Your Cheat Sheet for AI Engineering Interview – Questions and Answers.
A Markdown-only list of AI engineering interview questions, running from transformer internals through RAG, agents, fine-tuning, evaluation, and serving. It is aimed at people preparing for AI engineer, LLM engineer, and LLMOps interviews, not at anyone learning the material for the first time.
The agent and production questions are current. Context compaction, loop termination with token budgets, MCP versus function calling, and RLVR versus learned reward models are topics that older interview lists tend to skip. Many questions are framed as incidents, such as 'Your AI agent deleted a production database. How do you prevent irreversible actions?', which tests judgment rather than recall. The ordering moves from attention and KV cache memory estimates through to serving, so the sections build on each other.
Most answers are links to the author's own blog, YouTube videos, and LinkedIn or X posts, and the blog links overwhelmingly point to outcomeschool.com. The repo has no answers of its own to check a candidate against, and the links are only as durable as the posts they point to. Several questions have no answer at all, such as 'Explain WordPiece and SentencePiece' and 'What is the difference between sparse and dense embeddings?'. Some entries are titles rather than questions, such as 'Context Engineering', 'Loop Engineering', and 'Small Language Models (SLMs)', which makes the list harder to practise from.