The free, open-source guide to AI engineer interview questions and answers for the newest and highest-paid roles in tech — Applied AI Engineer and Forward Deployed Engineer (FDE) — at OpenAI, Anthropic, Palantir, and top AI startups. Covers LLM, RAG, AI agents, evals, prompt engineering, fine-tuning, and AI system design interview questions, built on the real 2026 interview bar.
★ Star this repo if it helps — it's how others find it. Practice these interactively, free → · Contribute a real interview question →
The Applied AI Engineer and Forward Deployed Engineer roles are new, they pay $300K–$1M+, and postings grew ~8× in a year — yet there's almost no structured interview prep for them. Generic "software engineer interview questions" don't cover large language models (LLMs), retrieval-augmented generation (RAG), AI agents, evaluations, or the forward-deployed craft. This handbook does — as a free, open, continuously-updated reference.
It's the companion to DEPLOYED, a free gamified trainer with spaced-repetition flashcards, retrieval checks, and timed mock interviews.
| # | Topic | Covers |
|---|---|---|
| 01 | Applied AI & Forward Deployed Engineer — the role | what AAE/FDE is, vs ML/research, the interview loop per company |
| 02 | LLM interview questions | tokens, embeddings, context windows, temperature, transformers, training |
| 03 | Prompt engineering interview questions | few-shot, chain-of-thought, structured output, robust prompting |
| 04 | RAG interview questions | chunking, embeddings, vector search, hybrid + rerank, RAG evaluation |
| 05 | AI agents & tool use interview questions | function calling, ReAct, MCP, multi-agent, when NOT to use an agent |
| 06 | LLM evaluation (evals) interview questions | eval sets, metrics, LLM-as-judge, offline vs online |
| 07 | LLM production interview questions | cost, latency, caching, reliability, prompt injection, security |
| 08 | Fine-tuning interview questions | SFT, LoRA/QLoRA, RLHF vs DPO, distillation, when to fine-tune |
| 09 | AI system design interview questions | a framework + worked end-to-end designs |
| 10 | FDE behavioral & customer-facing | ambiguity, STAR, customer scenarios, the forward-deployed craft |
| 11 | Round-by-round playbook | the decomposition round, take-homes & video walkthroughs, AI-assisted & debugging rounds, the discovery-call simulation |
Plus: 📋 Interview Question Bank (real-style questions + strong answers) · ⚡ Rapid-Fire Flashcards.
- Skim a topic, then use its retrieval checks — recall the answer before revealing it. Producing the answer is the learning.
- Drill flashcards across topics (interleaving beats single-topic cramming).
- Simulate — go through the question bank out loud, on a timer, then self-grade.
- Practice interactively in the free trainer → website.
- Lead with evals — "how do you know your AI system actually works?" is the differentiator.
- Start with the customer, not the tech — clarify before you architect.
The most valuable thing you can add is a question you were actually asked — role, company, month/year, and the question. It keeps this current (these loops change fast). See CONTRIBUTING.md.
applied ai engineer interview · forward deployed engineer interview questions · ai engineer interview questions and answers · llm interview questions · rag interview questions · ai agents interview · mcp interview · prompt engineering interview · llm evaluation interview · fine-tuning llm interview · ai system design interview · openai interview · anthropic interview · palantir fde interview · machine learning engineer interview 2026
Content under CC BY 4.0 — use it, share it, adapt it, just credit this repo.