pallavi-shekhar/ai-engineering-interview-questions-company-wise
Public
Star 历史趋势
数据来源: GitHub API · 生成自 Stargazers.cn
README.md

AI Engineering Interview Questions Company Wise

AI Engineering Interview Questions Company Wise

AI Engineering Interview Questions Company Wise - Your Cheat Sheet For AI Engineering Interviews at Top AI Companies

Real interview questions asked in AI Engineering interviews at 35 companies, organized company by company, with answers linked wherever we have them.

These interview questions and answers are helpful for roles such as:

  • AI Engineer
  • Gen AI Engineer
  • LLM Engineer
  • Agentic AI Engineer
  • AI Agent Engineer
  • Machine Learning Engineer
  • Research Engineer
  • Applied Scientist
  • Forward Deployed Engineer
  • AI Solutions Architect
  • AI Platform Engineer
  • Applied AI Engineer
  • LLM Inference and Performance Engineer
  • MLOps Engineer
  • LLMOps Engineer

Table of Contents

Prepared and maintained by Outcome School

AI and Machine Learning Program by Outcome School: AI and Machine Learning Program

Follow Outcome School

Note: We will keep updating this with new questions and answers.

For topic-wise questions and answers, see AI Engineering Interview Questions and Answers.

How to use this

  • Questions are compiled from publicly reported interview experiences. Nothing here is confidential. Interview loops change constantly and vary by team, level, and region, so treat each company section as a map of what that company cares about, not a script of what you will be asked.
  • Start with Common Questions Asked Across Companies. These are the questions that recur across many companies. Each one is listed once, with the companies that ask it, so nothing is repeated in the company sections.
  • Then go to your target companies. Each company section has the roles it covers, the interview loop as publicly reported, and the company-specific questions grouped by topic.
  • Wherever we have an answer, it is linked right below the question. We will keep adding answers.

Common Questions Asked Across Companies

These questions come up in AI Engineering interviews at many companies. Each question is listed once here, with the companies where it (or a company-specific version of it) is asked. Work through these first.

LLM Internals and Architecture

Inference, Serving and GPU Performance

RAG and Retrieval

Agents and Tool Use

Fine-Tuning, Post-Training and Alignment

Evaluation and Observability

  • Design an LLM-as-judge evaluation. What are its known biases and how do you correct for them?
  • How do you build an eval set when there is no labelled ground truth and experts are expensive?
  • How do you detect and measure hallucinations in a production RAG system?
  • Design the regression gate that decides whether a prompt or model change ships.
  • Why do benchmark scores improve while users say the system got worse? Enumerate the reasons.
  • What is benchmark contamination and how do you guard against it?
  • What observability does a production LLM system need: traces, spans, costs, feedback?
  • How do you manage prompt versioning and rollbacks in production?
  • Design online evaluation: what do you log, what do you sample, and what do you A/B?
  • How would you evaluate an agent, as opposed to a single model response?

Safety, Security and Responsible AI

  • What is prompt injection (direct and indirect), and what is your layered defence?
  • Walk me through the OWASP Top 10 for LLM applications and which ones actually bite in practice.
  • What is the difference between jailbreaking and adversarial prompting?
  • Design guardrails for a consumer-facing assistant. Input filters, output filters, or both?
  • What is Constitutional AI and how does it differ from RLHF? What is RLAIF?
  • How do you prevent an agent with tool access from exfiltrating data via a malicious web page?
  • How do you handle PII in prompts, logs and training data?
  • What is mechanistic interpretability and why do labs invest in it?
  • How would you audit a deployed model for differential performance across user groups?
  • Design a red-teaming programme for a model you are about to release.

Multimodal, Speech and Voice AI

AI System Design

Coding and Data Structures

Frontier AI Labs

Anthropic

Roles this covers: Member of Technical Staff (MTS), Software Engineer (product / infra / API-serving), Research Engineer, Research Scientist, Applied AI Engineer, Forward Deployed Engineer (Applied AI), Performance Engineer (inference & kernels), Product Engineer (Claude Code / Claude.ai).

Interview loop, as publicly reported: Recruiter screen (~30 min, substantive and failable) → CodeSignal-style or live coding assessment (~70–90 min, one practical problem in ~4 progressive levels) → virtual onsite of roughly five rounds: project deep-dive, one or two coding rounds, system design, and a dedicated values/culture round. Reported end-to-end: 3 weeks to ~2 months. Some MLE loops now include an AI-collaboration round where Claude is provided and you are graded on how you direct and verify it.

Also prepare: the common questions tagged Anthropic under LLM Internals and Architecture, Inference, Serving and GPU Performance, Agents and Tool Use, Evaluation and Observability, Safety, Security and Responsible AI, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • Build core business logic for a toy banking application: a spec that grows in four progressive levels against a black-box evaluator.
  • Build an in-memory database: SET/GET/DELETE first, then filtered scans, then TTL with timestamps, then file compaction.
  • Create a task scheduler.
  • Build an OOP system for managing courses, grades and students.
  • Given a helper method that crawls a URL, write a crawler over a domain: first synchronous, then make it async.
  • Convert nested stack traces into discrete start and end events.
  • Build a rate limiter. Every ten minutes I add a requirement: per-tenant limits, burst allowances, then a sliding window. How do you keep the code from collapsing?
  • You need to run an LLM call over 50,000 documents. The API allows ~100 concurrent requests and occasionally returns 429s and timeouts. Write the Python.
  • How would you parallelise this task? (Concurrency and data mutation come up repeatedly across rounds.)
  • SQL: write a query to find the top five pairs of products most frequently purchased together.
  • SQL: determine whether any user has overlapping subscription date ranges.
  • SQL: return each employee's current salary after an ETL error inserted a new salary row every year.

LLM Internals and Architecture

Inference, Serving and GPU Performance

Agents and Tool Use

Fine-Tuning, Post-Training and Alignment

  • Explain Constitutional AI. What does it buy you over vanilla RLHF, and what doesn't it solve?
  • How do scaling laws influence the safety evaluation of large models?

AI System Design

  • Design the Claude chat service.
  • Design a system that enables a large language model to handle multiple questions in a single thread.
  • Design a distributed search system for 1 billion documents at 1 million QPS.
  • Design APIs for developers to access Anthropic's models securely and efficiently.
  • Design a file-sharing / distribution system.

Evaluation and Observability

  • How would you design an experiment to test for a specific emergent capability or bias in a large language model?

Safety, Security and Responsible AI

  • Your agent reads inbound email and can send replies and search internal docs. Walk me through the prompt-injection attack surface and your defences.
  • What do you see as the most pressing unsolved problem in AI alignment?
  • How would you balance performance optimisation with model interpretability?
  • How would you approach designing a system to ensure the safe deployment of AI models in production?

Applied and Forward-Deployed Scenarios

  • An enterprise customer says “Claude hallucinates too much” in their RAG-based knowledge assistant. You're the applied engineer on the account. What happens in the first 48 hours?
  • How would you make complex AI research findings accessible to a non-technical audience?

Behavioral and Culture

  • Walk me through a project you owned end to end. What were the key technical decisions?
  • Why Anthropic specifically, and where do you disagree with Anthropic?
  • Tell me about a technical misjudgement that delayed a project.
  • What are your thoughts on AI safety and the risks of advanced AI systems?

OpenAI

Roles this covers: Member of Technical Staff, Software Engineer, Machine Learning Engineer, Research Engineer, Research Scientist, Applied AI Engineer, Forward Deployed Engineer, Solutions Architect, Data Scientist.

Interview loop, as publicly reported: Recruiter screen → technical screen (practical coding, often a build-something-real task rather than LeetCode) → onsite: two or three coding/practical rounds, a domain-depth or ML round, a system design round, and behavioural/mission-fit. Applied AI and FDE loops add a customer-scenario and solution-design round.

Also prepare: the common questions tagged OpenAI under LLM Internals and Architecture, Inference, Serving and GPU Performance, Agents and Tool Use, Fine-Tuning, Post-Training and Alignment, Evaluation and Observability, Safety, Security and Responsible AI, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • Design and implement an in-memory key-value store supporting set, transactional begin, commit and abort.
  • Create a database ORM, step by step.
  • Code a trivial web crawler using Go.
  • Implement a UI from a mockup with provided CSS and API.
  • Refactor bad code: here are ~120 lines of working but messy code with passing tests. Improve the architecture without breaking them. What do you change first?
  • Write a Python function that displays the first n Fibonacci numbers.
  • Infection-spread simulation.

ML and DL Fundamentals

  • Compute the KL divergence given different random variables.
  • If the accuracy of a classifier is 1, what is the lower/upper bound on the loss function for a single training example?
  • We have two models, 85% and 82% accuracy. Which do you pick?
  • How do you handle missing data in Pandas?

LLM Internals and Architecture

Agents and Tool Use

AI System Design

  • How would you build an LLM-powered enterprise search system?
  • Design the serving stack for a ChatGPT-scale consumer assistant: hundreds of millions of weekly users, streaming chat, multiple model tiers.
  • Design and build a webhook delivery system that reliably delivers events to customer-registered URLs.
  • Design a system to schedule jobs in a distributed environment.
  • Design an in-memory database. / Design Slack.

Evaluation and Observability

  • A customer says “the model got worse” after you upgraded model versions in their deployment. How do you verify and respond?
  • An enterprise customer reports that responses from your deployed system have gotten slow. Walk me through the diagnosis.

Safety, Security and Responsible AI

  • How do you approach GenAI safety in consumer products?
  • How would you design safeguards for an AI system that can take actions on behalf of a user?

Applied and Forward-Deployed Scenarios

  • An enterprise customer says: “We want AI to automate our claims processing.” You're the engineer in the room. What do the first two weeks look like?
  • Do you have experience working with APIs? Are you used to working with C-suite executives?

Behavioral and Culture

  • What is your favourite product and why?
  • Tell me about a time you made a mistake.
  • Tell me about a time you had a conflict with someone. How did you resolve it and what did you learn?
  • Tell me about a time you had conflicting priorities with stakeholders and how you secured alignment.
  • What is the project you are most proud of?

Google DeepMind and Google AI

Roles this covers: Research Engineer, Research Scientist, Machine Learning Engineer, Software Engineer (ML), Forward Deployed Engineer, Applied AI Engineer (Google Cloud / Vertex AI), Data Scientist.

Interview loop, as publicly reported: Recruiter screen → technical phone screen (coding, sometimes ML fundamentals) → onsite: two coding rounds, an ML domain/breadth round, an ML system design round, and Googleyness/leadership. DeepMind Research Engineer loops add a research deep-dive plus maths/probability and a from-scratch implementation round; hiring committee and team matching follow.

Also prepare: the common questions tagged Google DeepMind under LLM Internals and Architecture, Inference, Serving and GPU Performance, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • You are receiving an unbounded stream of event IDs. Return the k most frequent IDs seen so far, at any point, with bounded memory.
  • Closest key: given a dictionary with letter keys and lists of letters as values, find the closest key.
  • Write a function to compute root-mean-square error given y_pred and y_true lists.
  • Parse bigrams: extract two-word phrases from strings for NLP feature engineering.

ML and DL Fundamentals

  • Define the bias-variance trade-off and discuss the relationship between the two.
  • What are the assumptions of linear regression?
  • Distinguish regularization from validation: when is each the right tool?
  • Derive the gradient of cross-entropy loss with softmax inputs, and explain why we fuse them numerically.
  • Explain the SVD and give two places it shows up in modern deep learning.
  • On average, how many fair coin flips until you see two heads in a row? Walk me through it.
  • When would you choose Q-learning over policy gradients, and vice versa?
  • You have a binary loan-approval classifier and limited access to feature weights. How do you explain a rejection?

Fine-Tuning, Post-Training and Alignment

  • Your pretraining loss suddenly diverges at step 300k of a long run. Diagnose and fix it.
  • Design the training setup for a model that doesn't fit on one accelerator, say 70B parameters on a pod.

AI System Design

  • Design the serving system for a multimodal assistant (text + image in, streaming text out) at hundreds of millions of users.
  • Design a personalised recommendation system for rental listings using demographics, property metadata, amenities, price, reviews and location.
  • Design a classifier that predicts the optimal moment to insert a commercial break in a video.
  • How would you improve product search results, focusing on the fraction of relevant documents retrieved (recall)?
  • Justify using a neural network for a given problem: what do you need to know about the network, dataset, timeline and business context?

Evaluation and Observability

  • Build the evaluation harness for a new frontier model release. What does it need to do?
  • Do 1 million Seattle ride trips suffice to build an accurate ETA prediction model? How would you decide?

Behavioral and Culture

  • Tell me about a time you disagreed with a researcher or tech lead about priorities, and what happened.

Meta (Superintelligence Labs, FAIR, Llama)

Roles this covers: Machine Learning Engineer (E4–E7), Research Engineer, Research Scientist, AI Infrastructure Engineer, Software Engineer (ML), Applied Research Scientist.

Interview loop, as publicly reported: Recruiter screen → technical screen (2 coding problems in 45 min) → onsite: two coding rounds, one ML system design round, one ML domain/breadth round, and a behavioural (“Jedi”) round. Some 2026 loops now include an AI-assisted coding round in three stages: explore and fix issues, implement new functionality, extend and improve the system.

Also prepare: the common questions tagged Meta under LLM Internals and Architecture, Inference, Serving and GPU Performance, Multimodal, Speech and Voice AI, AI System Design.

Coding and Data Structures

  • Given an array nums of n integers where n > 1, return an output array (product of array except self).
  • Find the minimum window in S which will contain all the characters in T.
  • Serialize and deserialize a binary tree.
  • Convert a binary tree to a circular doubly linked list.
  • Alien dictionary: determine character ordering from a sorted word list.
  • K closest points to origin; top-k frequent elements; minimum number of conference rooms.
  • Regular expression matching with '.' and '*'.
  • Two-part warm-up: given a stream of user actions, return the k most engaged-with items. Then: why might your heap solution be the wrong choice in production?

ML and DL Fundamentals

  • Your ads CTR model shows a 2% offline AUC gain, but the online A/B is revenue-neutral with worse calibration. What is going on, and what do you do?

LLM Internals and Architecture

Inference, Serving and GPU Performance

  • You need to serve a Llama-class 70B+ model to hundreds of millions of assistant users. What does the serving stack look like and where does the money go?

Agents and Tool Use

  • You're dropped into an unfamiliar multi-file codebase with a failing behaviour and an LLM assistant available. Walk me through how you'd fix it.

Fine-Tuning, Post-Training and Alignment

AI System Design

  • Design the recommendation system for Instagram Reels.
  • Design a personalised news-feed ranking system / the “next post” logic for Facebook's feed.
  • Design a recommendation system for Facebook Ads, and an evaluation framework for ads ranking.
  • Design the ML components behind an Instagram Story feature.
  • Design an end-to-end classification pipeline for Marketplace listings.
  • Design a language translation model / service.

Evaluation and Observability

  • How would you build the evaluation system for a Meta AI assistant before and after each model release?

Safety, Security and Responsible AI

  • Design the harmful-content detection system for Facebook and Instagram uploads.

Multimodal, Speech and Voice AI

  • How do modern multimodal models get image and video understanding into an LLM, and what changes for video specifically?

Behavioral and Culture

  • Give me an example of a project where you used data and machine learning. What obstacles did you hit?
  • Tell me about a time you drove a significant result through ambiguity, and a time you were wrong.
  • Tell me about maintaining a production ML pipeline. Why Meta?

xAI

Roles this covers: Member of Technical Staff, AI Engineer, Infrastructure Engineer, Research Engineer, Product Engineer (Grok), Data / RL environments engineer.

Interview loop, as publicly reported: Fast, low-process loops. Typically a recruiter or hiring-manager screen → one or two live coding rounds heavy on data structures and systems → a practical build-something round (often a timed four-hour product build or a stubbed-out module in a real codebase) → founder/leadership conversation. Speed and raw shipping ability are weighted heavily.

Also prepare: the common questions tagged xAI under LLM Internals and Architecture, Inference, Serving and GPU Performance, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • Build an in-memory key-value store with SET/GET/DELETE, then add transactions with BEGIN/COMMIT/ROLLBACK, including nested transactions.
  • Write an iterator class that lazily flattens an arbitrarily nested list of lists/integers: no generators, explicit state.
  • Here is a scheduler class from a small LLM inference engine. One method, _admit_requests, is a stub: no spec, no docstring, no tests. Walk me through your first thirty minutes.

Inference, Serving and GPU Performance

  • Estimate the KV-cache memory to serve a 70B-class model at 128k context. What do you do when it doesn't fit?
  • Design a rate limiter for an LLM API where cost scales with tokens, not requests.

Fine-Tuning, Post-Training and Alignment

  • You're training on tens of thousands of GPUs and hardware fails constantly. How do you keep goodput high?
  • Loss spikes mid-run on a large pretraining job. Walk me through your debugging process.
  • Design a deduplication pipeline for a web-scale pretraining corpus. It has to run as a streaming process.

AI System Design

  • Design the serving stack for a consumer chatbot with real-time search over a social-media firehose.

Behavioral and Culture

  • You have four hours to build and demo a working AI-powered product. How do you spend them?

Mistral AI

Roles this covers: Research Engineer, ML Engineer, Applied AI Engineer, Solutions Architect / Forward Deployed Engineer, Inference Engineer, Platform Engineer.

Interview loop, as publicly reported: Recruiter screen → technical screen (Python + ML fundamentals) → pair-programming round building a small LLM-backed service → deep-dive on transformer/serving internals → customer-scenario round for applied roles → culture/founders round. European enterprise and on-prem deployment context shows up throughout.

Also prepare: the common questions tagged Mistral AI under LLM Internals and Architecture, Inference, Serving and GPU Performance, Agents and Tool Use, Fine-Tuning, Post-Training and Alignment, Coding and Data Structures.

Coding and Data Structures

  • Pair-programming: build a service that takes a user question, enriches it with data from a third-party API, and answers via a chat-model API. How do you structure it?

LLM Internals and Architecture

Inference, Serving and GPU Performance

Agents and Tool Use

Fine-Tuning, Post-Training and Alignment

  • After fine-tuning on a customer's task, target accuracy is up but the model got worse at everything else. What happened and what do you do?

AI System Design

  • Design an on-prem deployment of an open-weight model for a European bank that cannot send data to any external API.

Cohere

Roles this covers: Member of Technical Staff, ML Engineer, Applied AI Engineer, Solutions Architect / Forward Deployed Engineer, Platform & Inference Engineer.

Interview loop, as publicly reported: Recruiter screen → technical screen (practical Python, streaming/API-shaped problems) → onsite: coding, retrieval/RAG depth, enterprise deployment design, customer-scenario round, and a values round. Remote-first; autonomy and ownership are explicitly tested.

Also prepare: the common questions tagged Cohere under LLM Internals and Architecture, RAG and Retrieval, Fine-Tuning, Post-Training and Alignment, Evaluation and Observability, Coding and Data Structures.

Coding and Data Structures

  • Design a token-based rate limiter for a multi-tenant LLM API. Implement the core, then tell me what changes when it's distributed.

LLM Internals and Architecture

  • Our flagship is a sparse MoE with ~10x more total than active parameters. Why is that architecture a good fit for private enterprise deployment, and where does it hurt?

RAG and Retrieval

  • You have an embedding model and a reranker. Why sell both? Design the two-stage retrieval pipeline and tell me when the reranker earns its latency.
  • An enterprise wants semantic search over ~100M documents but is balking at vector-index cost. Walk me through embedding compression options and the maths.
  • How would you evaluate multilingual retrieval quality when employees query in French and Korean over mostly-English documents?
  • A customer 10x'd their indexed documents and reports answer quality “got noticeably worse.” Drive the investigation.

Agents and Tool Use

  • Design an agent that automates an enterprise workflow, say, drafting RFP responses from internal documents and a CRM. What does “enter-prise-grade” add?

AI System Design

  • A bank wants the whole stack (model, RAG, agents) deployed air-gapped on their own GPUs. What actually changes versus your SaaS?

Evaluation and Observability

  • An enterprise customer wants to deploy your RAG system but has no labelled data. How do you evaluate it before and after launch?

Behavioral and Culture

  • Tell me about a time you owned an ambiguous problem end-to-end without much direction.

DeepSeek

Roles this covers: Research Engineer, Infrastructure / Systems Engineer, Inference Engineer, Data Engineer, Algorithm Engineer.

Interview loop, as publicly reported: Heavily research- and systems-weighted: paper deep-dive, from-scratch implementation rounds in PyTorch, distributed-training and low-precision depth, plus an algorithmic coding round. Expect direct questions about their published architecture and training papers.

Also prepare: the common questions tagged DeepSeek under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment.

LLM Internals and Architecture

Inference, Serving and GPU Performance

  • Sketch how you would serve a 671B-parameter MoE model with low latency under GPU-memory constraints.

Fine-Tuning, Post-Training and Alignment

  • R1-Zero was trained with RL and essentially no SFT first. What did that show, and why did full R1 add SFT back?
  • FP8 training at 671B scale is hard. What actually breaks in low precision, and how do you make it stable?
  • How do you build a training dataset without triggering model collapse when much of your data is synthetic?
  • DualPipe overlaps computation and communication in training. Why is that overlap the whole game at this scale, and what is the trade-off?

Behavioral and Culture

  • DeepSeek claims frontier-class results at a fraction of the usual training cost. If an interviewer asks “how is that even possible,” what is your structured answer?

Moonshot AI (Kimi)

Roles this covers: Research Engineer, Infrastructure Engineer, Inference / Serving Engineer, Agent Engineer.

Interview loop, as publicly reported: Research and long-context systems focused: architecture deep-dive, distributed serving design, a PyTorch implementation round, and an agentic-evaluation discussion.

Also prepare: the common questions tagged Moonshot AI under LLM Internals and Architecture, Inference, Serving and GPU Performance, Evaluation and Observability, Coding and Data Structures.

LLM Internals and Architecture

Inference, Serving and GPU Performance

  • Walk me through why you would disaggregate prefill and decode onto separate machines, as Mooncake does. What does that buy you and what does it cost?
  • A chat assistant re-sends a long conversation history on every turn. How do you avoid recomputing all of it, and what are the pitfalls?

RAG and Retrieval

Fine-Tuning, Post-Training and Alignment

  • Training a trillion-parameter model, attention logits can blow up and destabilise the run. What is going on, and how does something like MuonClip address it?
  • Kimi K1.5 scaled RL for reasoning without a process reward model or tree search. Why deliberately keep the RL recipe that simple?

Evaluation and Observability

  • Kimi K2 targets agentic and coding tasks. How would you evaluate whether an agentic model is actually good, beyond a single benchmark number?

Zhipu AI (GLM)

Roles this covers: Research Engineer, Agent Engineer, RL Infrastructure Engineer, Applied AI Engineer.

Interview loop, as publicly reported: Architecture and post-training depth, an RL-infrastructure design round, a GUI-agent design round, and an implementation round in PyTorch.

Also prepare: the common questions tagged Zhipu AI under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment, Evaluation and Observability.

LLM Internals and Architecture

  • GLM's original pre-training objective is autoregressive blank infilling. How does it differ from BERT and GPT, and why did the team argue it unifies understanding and generation?
  • GLM-4.5 is an MoE with 355B total but 32B active parameters. Explain the economics: what does that split buy you and what does it cost?
  • Implement a top-k MoE router in PyTorch. Then contrast auxiliary-loss load balancing with a loss-free approach.
  • What is Multi-Token Prediction (MTP), why add an MTP layer, and how does it help at inference time?
  • GLM has been bilingual Chinese/English since GLM-130B. What changes in tokenization, data and evaluation when a model must serve both languages well?

Agents and Tool Use

  • AutoGLM and CogAgent operate real GUIs from screenshots over tens of steps. Design the agent: perception, action space, and error recovery for a 50-step task.

Fine-Tuning, Post-Training and Alignment

AI System Design

  • Design AutoGLM end to end: a cloud service letting users delegate multi-step phone tasks (“order my usual coffee”) to an autonomous agent. Architecture and failure modes.

Evaluation and Observability

  • How would you evaluate an agentic coding model on SWE-bench and τ-bench style benchmarks without fooling yourself?

Alibaba (Qwen)

Roles this covers: Algorithm Engineer (LLM), Research Engineer, Inference Engineer, Multimodal Engineer, Applied AI Engineer (Alibaba Cloud / Model Studio).

Interview loop, as publicly reported: Classic Alibaba structure: two or three technical rounds (algorithms + ML depth), a cross-examination round with a senior manager, and an HR round, with Qwen-specific architecture and multilingual questions layered on top.

Also prepare: the common questions tagged Alibaba under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment, Evaluation and Observability, Multimodal, Speech and Voice AI, Coding and Data Structures.

Coding and Data Structures

  • Qwen2.5-Coder trains with repository-level fill-in-the-middle using tokens like <|fim_prefix|>, <|fim_suffix|>, <|repo_name|>. Write the function that formats a repo-level FIM example, and explain why repo-level beats file-level.

LLM Internals and Architecture

  • Qwen uses byte-level BPE with a ~151K vocabulary, augmented for multilingual coverage and with digits split into single characters. Why those choices, and what are the trade-offs?
  • Qwen3 unifies a thinking mode and a non-thinking mode in one model with a caller-settable thinking budget. How would you train that, and how would you serve it?
  • Qwen ships both dense and MoE models (30B with ~3B active; 235B with ~22B active). When would you pick the 30B-A3B MoE over a 32B dense?
  • Qwen2.5 extends context to 128K (and ~1M for Turbo) using YaRN plus Dual Chunk Attention, mostly training-free. Explain how, and why post-hoc extension is attractive.

Fine-Tuning, Post-Training and Alignment

Evaluation and Observability

  • Qwen ships open weights that top public leaderboards. As the release engineer, how do you make sure the benchmark numbers are trustworthy and not contaminated?

Multimodal, Speech and Voice AI

  • Qwen2.5-VL uses a native dynamic-resolution ViT with window attention and multimodal RoPE. Why native resolution instead of fixed tiling, and what does MRoPE encode?

Behavioral and Culture

  • Alibaba open-sources Qwen under Apache 2.0 while running a commercial cloud business. Walk me through the strategy, and tell me about an ambiguous technical decision you owned end to end.

Sarvam AI

Roles this covers: Research Engineer (LLM / speech), ML Engineer, Applied AI / Forward Deployed Engineer, Speech Engineer, Edge / Inference Engineer.

Interview loop, as publicly reported: Research and applied rounds side by side: Indic NLP and tokenizer depth, a speech/ASR round, an implementation round, a deployment-on-constrained-hardware round, and a government/enterprise deployment scenario.

Also prepare: the common questions tagged Sarvam AI under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment, Multimodal, Speech and Voice AI.

Coding and Data Structures

  • Write code to measure a tokenizer's fertility across languages, and explain what you would do with the result.

LLM Internals and Architecture

Inference, Serving and GPU Performance

RAG and Retrieval

  • Design cross-lingual RAG: the knowledge base is in English and Hindi, but users ask in Tamil, Telugu or transliterated Hinglish.

Fine-Tuning, Post-Training and Alignment

AI System Design

  • Design a real-time voice agent for a citizen helpline in Hindi and three regional languages, targeting sub-250 ms perceived latency over a phone line.

Evaluation and Observability

  • How would you evaluate an Indic LLM properly? Why is running translated English benchmarks not enough?

Multimodal, Speech and Voice AI

  • Build a Voice Activity Detector from scratch. How do you make it robust for phone-quality Indian-language audio?
  • Whisper transcribes Hinglish poorly, often forcing output into one language or hallucinating. Why, and how would you build an ASR that handles code-mixed speech?
  • Bulbul-style TTS has to speak code-mixed, mixed-script text naturally. What are the hard parts of text normalization and prosody for Indian-language TTS?

Applied and Forward-Deployed Scenarios

  • A state agency wants to move a paper-and-call-centre welfare-scheme service onto a multilingual assistant, on-prem for data residency. How do you scope and ship it?

Big Tech AI Organizations

Microsoft

Roles this covers: AI Engineer, Applied Scientist, Machine Learning Engineer, Software Engineer (AI Platform / Copilot), Azure AI Solutions Architect, Principal Applied AI Engineer.

Interview loop, as publicly reported: Recruiter screen → technical phone screen (coding + a little ML) → onsite loop of 4–5 rounds: two coding, one ML/AI depth, one AI system design or low-level design, and an as-appropriate-hire round with a senior leader. Azure AI and Copilot roles add a customer-architecture round.

Also prepare: the common questions tagged Microsoft under Inference, Serving and GPU Performance, RAG and Retrieval, Agents and Tool Use, Fine-Tuning, Post-Training and Alignment, Safety, Security and Responsible AI, AI System Design.

Coding and Data Structures

  • Implement “top-k most frequent search queries” over a large query log, then tell me what breaks when the log becomes an unbounded stream across many machines.
  • Low-level design: sketch the classes and interfaces for the tool-calling layer of an agent host, where tools can come from native code, an OpenAPI spec, or an MCP server.

Inference, Serving and GPU Performance

AI System Design

  • Design a Copilot feature that answers questions over a user's work email, documents and meetings, without ever leaking content the user can't access.
  • Design an agent that can take actions in a spreadsheet (“insert a pivot table of Q3 sales by region”): orchestration, tools and failure handling.

Evaluation and Observability

  • How would you evaluate a meeting-summarisation feature before shipping it to a hundred million users?

Safety, Security and Responsible AI

  • Your Copilot summarises incoming email. An attacker emails a target user with hidden instructions addressed to the model. Walk me through the attack and your defence.
  • A shipped Copilot feature that summarises job applicants for recruiters is accused of working worse for some groups. How do you establish whether that's true, and what do you do about it?

Behavioral and Culture

  • Tell me about a time a technical decision you championed turned out to be wrong. What happened, and what did you change afterward?

Amazon (AWS)

Roles this covers: Applied Scientist (I/II/III), Machine Learning Engineer, Data Scientist, Software Development Engineer (AI/ML), GenAI Specialist Solutions Architect, Applied AI Engineer (Bedrock, Q, SageMaker).

Interview loop, as publicly reported: Online assessment or phone screen → onsite “loop” of 4–6 rounds, each anchored on Leadership Principles: two coding, one ML breadth, one ML depth / research deep-dive or ML system design, one bar-raiser. Applied Scientist loops add a research presentation.

Also prepare: the common questions tagged Amazon under LLM Internals and Architecture, Inference, Serving and GPU Performance, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • Find the top-K most frequent items in a high-volume event stream with bounded memory.
  • Divide two integers without using multiplication, division or modulo. Find the number of connected components in a graph. Check balanced parentheses.

ML and DL Fundamentals

LLM Internals and Architecture

Inference, Serving and GPU Performance

Agents and Tool Use

  • Design an agent that operates a web browser to complete multi-step tasks. How do you make it reliable enough to ship?

AI System Design

  • Design a multi-tenant inference platform that serves many foundation models to thousands of customers (Bedrock-shaped).
  • How would you design a recommendation system to suggest books to users? How would you model a warehouse inventory problem?

Evaluation and Observability

  • How would you decide an LLM-powered assistant is ready to launch to millions of customers?

Behavioral and Culture

  • Tell me about a time you disagreed with your team's technical direction. What did you do? (Have Backbone; Disagree and Commit)
  • Tell me about your most significant failure. What happened, and what did you change afterward?
  • Tell me about a time you saw an opportunity to do something bigger than the initial scope. (Think Big)

Apple

Roles this covers: Machine Learning Engineer, AI/ML Research Engineer, On-device ML Engineer, Software Engineer (Apple Intelligence / Siri), Applied Scientist.

Interview loop, as publicly reported: Recruiter screen → hiring-manager technical call → onsite of 4–6 rounds with the specific team: coding, ML depth, on-device/efficiency depth, system design, and behavioural. Secrecy means you may be interviewed for work you cannot be told about.

Also prepare: the common questions tagged Apple under LLM Internals and Architecture, Inference, Serving and GPU Performance, Agents and Tool Use, Fine-Tuning, Post-Training and Alignment, Coding and Data Structures.

Inference, Serving and GPU Performance

Agents and Tool Use

Fine-Tuning, Post-Training and Alignment

  • You have one on-device base model but a dozen features: summarization, rewriting, reply suggestions, tone adjustment. How do you specialise without shipping a dozen models?
  • How would you improve an on-device model using signals from user devices without collecting user content?

AI System Design

Evaluation and Observability

  • You're shipping notification summarization to hundreds of millions of users in 30+ locales, and you cannot log user content. Design the evaluation and regression-detection story.

Behavioral and Culture

  • Tell me about a time you had to make progress with incomplete information: you couldn't be told the full context of what you were building.

NVIDIA

Roles this covers: Deep Learning Software Engineer (Inference / LLM Performance), CUDA Kernel Engineer, Machine Learning Engineer, Solutions Architect, Applied Scientist, TensorRT-LLM / Dynamo engineer.

Interview loop, as publicly reported: Recruiter screen → hiring-manager technical call → onsite of 4–6 rounds: CUDA/C++ or Python coding, GPU performance and roofline reasoning, LLM inference depth, ML fundamentals, and system/solution design. Solutions Architect loops replace kernel depth with customer-scenario rounds.

Also prepare: the common questions tagged NVIDIA under LLM Internals and Architecture, Inference, Serving and GPU Performance.

Coding and Data Structures

  • Here's a CUDA kernel that's 10x slower than expected. Without running it, what are the usual suspects, and how do you confirm each?
  • Implement the block manager for a paged KV cache: allocate, append, free, and copy-on-write prefix sharing.
  • A model runs fine in FP32 but produces garbage after conversion to FP16. Debug it.

ML and DL Fundamentals

  • Explain strategies to combat overfitting in tree-based classification models.
  • Summarize the differences and benefits of the Adam optimizer compared with other methods for neural-network image classification.
  • A network confuses pugs and pit bulls and some training labels are wrong. How do you modify the model and the data?
  • How do you evaluate a clustering model's effectiveness without pre-labelled groups?

Inference, Serving and GPU Performance

  • You want to serve a 70B-parameter model on a single 80 GB GPU. Walk me through whether it fits and what single-stream tokens/sec you'd expect.
  • What does TensorRT / TensorRT-LLM actually do to a model to make it faster, and when will it not help?
  • Design the parallelism strategy for serving a 405B-parameter dense model. TP, PP, EP: what goes where and why?

AI System Design

  • Design a podcast search engine with transcript indexing. / Design a recommendation algorithm for type-ahead search.

Applied and Forward-Deployed Scenarios

Behavioral and Culture

  • Describe a time you dealt with conflicting priorities or stakeholder feedback. What would your current manager say about you?

Tesla

Roles this covers: AI / ML Engineer (Autopilot, Optimus), Deep Learning Engineer, Computer Vision Engineer, Data Engineer (Autopilot), Inference / Silicon software engineer.

Interview loop, as publicly reported: Recruiter screen → hiring-manager technical call → onsite: coding (often C++/Python), deep-learning depth focused on vision and training pipelines, a data/infra round, and a hands-on debugging round. Loops are fast and heavy on practical engineering.

Coding and Data Structures

  • Implement non-maximum suppression. Then vectorise it.
  • Write an efficient ring buffer for high-rate sensor data with a fixed memory budget.

ML and DL Fundamentals

  • How would you design the neural network architecture for multi-camera 3D object detection without lidar?
  • How do you handle extreme class imbalance in rare-event detection (e.g. a child running into the road)?
  • Explain how you would auto-label a fleet dataset and what quality controls you would put on it.
  • How would you detect and handle distribution shift between fleet data and your training set?

Inference, Serving and GPU Performance

  • The onboard compute budget is fixed. Walk me through quantizing and pruning a vision model without losing recall on small objects.

AI System Design

  • Design the data engine: fleet triggers → upload → labelling → retraining → shadow-mode validation → release.

Evaluation and Observability

  • Disengagement rate is a weak proxy. How would you actually measure whether an autonomy release is safer than the last one?

Multimodal, Speech and Voice AI

  • How would you fuse camera, radar and IMU inputs into a single perception stack, and where would you fuse them?

Behavioral and Culture

  • Tell me about the most technically demanding thing you have shipped, and what you would do differently.

Consumer-Scale ML Companies (Uber, Netflix, LinkedIn, Airbnb, Pinterest, Spotify)

Roles this covers: Machine Learning Engineer, Senior/Staff MLE, Applied Scientist, ML Platform Engineer, GenAI Engineer.

Interview loop, as publicly reported: A consistent shape across all of them: coding screen → onsite with two coding rounds, one ML system design round (the differentiator), one ML breadth/depth round, and behavioural. GenAI rounds have been added to most of these loops since 2024.

Also prepare: the common questions tagged Consumer-Scale ML Companies under AI System Design.

Coding and Data Structures

  • Implement a streaming top-k with a bounded-memory sketch; implement a sliding-window rate counter.

ML and DL Fundamentals

  • Your offline metric improved but the online A/B did not. Enumerate the reasons this happens and how you would tell them apart.
  • Explain position bias in ranking data and how you would debias training.
  • How do you design a feature store, and what causes training/serving skew?

AI System Design

  • Design the ETA prediction system for a ride-hailing marketplace. What features, what model, how do you serve it in <100 ms?
  • Design a personalised feed ranking system with a two-stage candidate generation and ranking architecture.
  • Design a content recommendation system for a streaming catalogue, including cold-start for new titles and new users.
  • Design “people you may know” / job-recommendation ranking at a professional network's scale.
  • Design dynamic pricing / surge for a two-sided marketplace and describe the feedback loops that can go wrong.
  • Design a visual search system: user uploads an image, you return visually similar in-catalogue items.
  • Design a fraud-detection system with heavy class imbalance and an adversarial opponent.
  • Design an LLM-powered customer-support assistant on top of an existing help centre, with escalation to humans.

Evaluation and Observability

  • How do you monitor a deployed ranking model for drift, and what triggers a retrain?

Behavioral and Culture

  • Tell me about a model you shipped that made a measurable business difference, and one that did not.

AI Infrastructure and Platform Companies

Databricks

Roles this covers: Software Engineer (ML Platform / Mosaic AI), Machine Learning Engineer, GenAI Solutions Architect, Forward Deployed / Delivery Solutions Architect, Applied AI Engineer.

Interview loop, as publicly reported: Recruiter screen → technical screen (practical coding, often concurrency or data-heavy) → onsite: two coding rounds, a distributed-systems or Spark-internals round, a GenAI/ML design round, and a customer-scenario round for field roles.

Also prepare: the common questions tagged Databricks under Inference, Serving and GPU Performance, RAG and Retrieval, Fine-Tuning, Post-Training and Alignment, AI System Design.

Coding and Data Structures

  • Implement a thread-safe batching logger: many producer threads call log(msg); a background thread flushes batches of up to 100 messages every second or when full.
  • You have a stream of billions of events and need the top-K most frequent keys with bounded memory. Exact is impossible: what do you do?
  • Given allowed IP ranges as CIDR blocks plus explicit deny ranges, implement is_allowed(ip) efficiently for millions of checks per second.
  • A Spark job joining a 2 TB fact table to a 50 GB dimension table has one straggler task running 100x longer than the rest. Diagnose and fix it.
  • A Structured Streaming job reads Kafka and writes to a Delta table. The cluster is killed mid-batch and restarts. Does the customer get duplicate rows? Explain at the level of the checkpoint and the transaction log.

Fine-Tuning, Post-Training and Alignment

Evaluation and Observability

  • Take a working GenAI agent prototype to production for an enterprise. What's your checklist between demo and launch?

Applied and Forward-Deployed Scenarios

  • A customer insists on fine-tuning an open model on their support tickets because “we want our own model.” You think RAG solves it. What do you do?
  • An agent you shipped four months ago runs on a base model being deprecated in 60 days. How do you swap the model without regressing quality, and what had to be in place beforehand?

Groq

Roles this covers: Compiler Engineer, Runtime / Systems Engineer, Inference Engineer, Silicon Software Engineer, Solutions Architect.

Interview loop, as publicly reported: Deep systems and compiler rounds: roofline and memory-hierarchy reasoning for an SRAM-only machine, compiler IR design, host-runtime design, a debugging round, and a customer/unit-economics round.

Also prepare: the common questions tagged Groq under Inference, Serving and GPU Performance.

Coding and Data Structures

  • Our compiler statically schedules every instruction and every chip-to-chip transfer. What does that compiler need to know that an NVCC-style compiler does not, and what breaks when it's wrong?
  • Design the IR and pass pipeline for a compiler targeting a spatial dataflow accelerator. Where does the memory-residency decision live, and why?
  • Write the host-side runtime that feeds a deterministic accelerator across many chips. What is genuinely hard about it?
  • A model passes bit-exact against the functional simulator on one chip but produces wrong output at rack scale. How do you find it?

Inference, Serving and GPU Performance

  • An LPU has no HBM at all, just on-die SRAM. Redo the decode roofline argument for that machine and tell me what changes.
  • A 70B dense model at 8-bit weights, chips with ~230 MB of SRAM each. Walk me through the deployment and the unit economics.
  • On a GPU you batch to amortise weight reads. What is the batching calculus on an SRAM-only machine, and how should that change how we price?
  • Determinism is the headline claim. What does it actually buy at p99, and why does it matter especially for agentic workloads?
  • How would you serve a large mixture-of-experts model on a statically scheduled fabric when expert selection is data-dependent?

AI System Design

Applied and Forward-Deployed Scenarios

  • A prospective customer runs their workload on H100s. Talk me through when you would tell them not to move.

Behavioral and Culture

  • Tell me about a performance optimisation you shipped. Give me the numbers, and tell me why I should believe them.

Together AI

Roles this covers: Inference Engineer, Kernel Engineer, ML Systems Engineer, Solutions / Forward Deployed Engineer, Platform Engineer.

Interview loop, as publicly reported: Inference-performance depth, a scheduler/serving design round, a practical streaming-server coding round, a distributed-training debugging round, and a customer-advisory round for field roles.

Also prepare: the common questions tagged Together AI under LLM Internals and Architecture, Inference, Serving and GPU Performance.

Coding and Data Structures

Inference, Serving and GPU Performance

Fine-Tuning, Post-Training and Alignment

  • A customer's distributed training job on your GPU cluster gets 55% scaling efficiency at 64 nodes. Debug it.

AI System Design

Applied and Forward-Deployed Scenarios

  • A customer wants to migrate from a proprietary frontier-model API to an open model. How do you run that engagement?

Hugging Face

Roles this covers: ML Engineer (open-source maintainer), Research Engineer, Infrastructure Engineer, Developer Advocate Engineer, Inference Engineer.

Interview loop, as publicly reported: Open-source-flavoured: a library-internals deep dive, a maintainer/code-review round, a practical fine-tuning or memory-budget round, a Hub/systems design round, and a culture round. Your public GitHub history is genuinely part of the evaluation.

Also prepare: the common questions tagged Hugging Face under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment.

Coding and Data Structures

  • transformers famously repeats code: each model gets its own self-contained modeling file. Defend that decision, then critique it.
  • Why did Hugging Face create safetensors when pickle-based checkpoints already worked everywhere?
  • A user loads a 2 TB dataset with datasets on a 64 GB RAM machine and it works. How? And when does it stop working?

LLM Internals and Architecture

  • Walk me through what actually happens when someone calls AutoModelForCausalLM.from_pretrained(…, device_map=“auto”, torch_dtype=“auto”).
  • Compare BPE, WordPiece and Unigram tokenization. Why is tokenizers written in Rust, and what tokenizer bugs bite people in practice?
  • What problem do chat templates solve, and what goes wrong when they're ignored?

Fine-Tuning, Post-Training and Alignment

  • Fine-tune an 8B model on a single 24 GB GPU. Walk me through the memory maths and the exact stack you'd use.
  • You're building a web-scale pretraining corpus (FineWeb-style). Walk me through the pipeline and how you decide whether each filter earns its place.

AI System Design

  • Design the Hugging Face Hub: millions of git repos where individual files are tens to hundreds of GB.
  • Design the serverless inference layer: any of thousands of Hub models can receive a request at any moment.

Behavioral and Culture

  • A community contributor opens a PR adding a new model architecture to transformers. You're the reviewing maintainer: what do you check, and how do you handle the interaction?

Scale AI

Roles this covers: Software Engineer, Machine Learning Engineer, Research Engineer (SEAL evals), Forward Deployed Engineer, Product Engineer (data engine, RL environments).

Interview loop, as publicly reported: Recruiter screen → technical screen (practical coding) → onsite: coding with progressive requirements, a data-quality/annotation design round, an evaluation-design round, and a customer-scenario round for FDE.

Also prepare: the common questions tagged Scale AI under LLM Internals and Architecture, Fine-Tuning, Post-Training and Alignment, Evaluation and Observability, AI System Design.

Coding and Data Structures

  • Build the task-lifecycle core of an annotation platform. Start simple; I'll add consensus of k annotators, then priority re-review, then annotator cooldowns.
  • Given annotation sessions as (start, end) timestamps, return the peak number of concurrent annotators and the intervals at peak load.

Fine-Tuning, Post-Training and Alignment

AI System Design

  • Design an end-to-end pipeline producing RLHF preference data for a frontier lab: 100k prompt-response comparisons a week, with quality guarantees.
  • Design a private LLM benchmark and leaderboard (SEAL-style). How do you keep it trustworthy as labs optimise against it?

Evaluation and Observability

  • Your annotators have no ground truth: the tasks are subjective preference judgments. How do you measure and improve label quality?
  • How would you benchmark an LLM agent's tool use, say, for enterprise workflows composing 10+ APIs?
  • An eval pipeline you own suddenly reports a 6-point drop for a customer's model between Tuesday and Wednesday. The model didn't change. Debug it.

Safety, Security and Responsible AI

  • Some annotators are pasting your tasks into ChatGPT and submitting the output. How do you detect and handle it?

Applied and Forward-Deployed Scenarios

  • An enterprise wants a document-Q&A assistant over 2M internal documents, pilot in four weeks, and their security team forbids data leaving their VPC. Scope and design it.
  • A robotics customer asks for 50,000 hours of manipulation demonstrations across 12 tasks and three embodiments. Design the collection pipeline, and tell me what makes one demonstration worth keeping.

Perplexity

Roles this covers: Software Engineer (search / infra), AI Engineer, Research Scientist, ML Engineer (ranking & retrieval), Product Engineer (Comet).

Interview loop, as publicly reported: Recruiter screen → technical screen → onsite: a system design round on retrieval/ranking at web scale, an ML/search deep-dive, a practical coding round, a product/craft round, and behavioural. Product taste about their own app is explicitly evaluated.

Also prepare: the common questions tagged Perplexity under LLM Internals and Architecture, Inference, Serving and GPU Performance, RAG and Retrieval, Evaluation and Observability, AI System Design, Coding and Data Structures.

Coding and Data Structures

  • Implement a client pool over multiple LLM providers with failover: providers fail, time out, or rate-limit, and callers should just get a completion.
  • You're ingesting millions of web pages a day. Detect near-duplicates (same article, different boilerplate) efficiently.
  • You need to embed millions of text chunks. The embedding service takes batches with a max batch size and a max total-token limit. Write the batcher and make it fast.
  • Implement beam search for an autoregressive model. When would an answer engine actually use it?

Inference, Serving and GPU Performance

RAG and Retrieval

AI System Design

  • Design an answer engine: a user types a question and gets a cited, streamed answer. Your end-to-end budget is 3 seconds to a complete short answer.
  • Design the retrieval pipeline pulling from 100B web pages with sub-second latency and freshness guarantees.
  • Design the ranking system combining BM25, dense retrieval and LLM reranking across multiple indexes.
  • Design Comet's hybrid browser architecture combining on-device privacy with cloud AI assistance.
  • How does an answer engine handle breaking news: a query about something that happened 20 minutes ago?

Evaluation and Observability

  • How would you evaluate answer quality for an answer engine, continuously and at scale, with both automated and human signals?

Safety, Security and Responsible AI

  • Design the citation-verification system to reduce hallucinations in generated answers. How do you ensure every claim is actually supported by its cited source?

Behavioral and Culture

  • What makes a Perplexity answer great vs mediocre? Where does Perplexity lose to traditional search, and where does it win? You clearly use it: what's broken, and what would you ship to fix it?

AI-Native Product Companies

Cursor (Anysphere)

Roles this covers: Software Engineer (product, infra, model serving), ML Engineer, Research Engineer, Infrastructure Engineer.

Interview loop, as publicly reported: Recruiter/hiring-manager screen (~45 min) → one to three 60-minute technical screens where you build data structures against Cursor's actual codebase (AI-tool access varies by round) → a two-day in-person project onsite (or an ~8-hour remote version) where you design and ship a feature on real Cursor code. Scoping, autonomy and effective AI-tool use are explicitly graded.

Also prepare: the common questions tagged Cursor under Inference, Serving and GPU Performance, RAG and Retrieval, Evaluation and Observability, AI System Design.

Coding and Data Structures

  • Build a hash tree to organise data in a repository.
  • Given a repository snapshot (path → content), build a Merkle tree and write the function returning which files changed between two snapshots without comparing every file's content.
  • Print the top view of nodes in a binary tree.
  • Find duplicate files in a file system.
  • Implement the core of an editor text buffer: efficient insert/delete at arbitrary positions and fast line lookup. What structure do you pick?

Inference, Serving and GPU Performance

RAG and Retrieval

Agents and Tool Use

  • Design the harness for an agent that makes multi-file changes from a natural-language task. How do you keep it from wrecking a codebase?
  • Design an agentic AI system that can autonomously adapt to new tasks.

AI System Design

  • Design Cursor's tab (next-edit prediction) system: it must feel instant (sub-100 ms perceived latency) for millions of daily users.
  • How would you index a 100k-file monorepo so an AI editor can retrieve relevant context, and keep the index fresh as the user edits?
  • The model is streaming a multi-file edit while the user keeps typing in one of those files. How do you apply the edits without corrupting the buffer?
  • Your agent model outputs an edited version of a 500-line file. Applying it verbatim is slow and error-prone. How do you make “apply” fast and reliable?
  • An agent needs to iterate on code (run builds, tests, lints) without disturbing what the user sees in their editor. Architect that.

Evaluation and Observability

  • How do you evaluate a code-editing model before shipping it? Design the offline and online eval story for tab or agent edits.

Behavioral and Culture

  • You have two days in our codebase and no assigned task. What do you build, and how do you spend the time?
  • Tell me about a time you made short-term sacrifices for long-term gains.

Cognition (Devin, Windsurf)

Roles this covers: Member of Technical Staff, Agent Engineer, Infrastructure Engineer, Deployed Engineer, Research Engineer (agent RL).

Interview loop, as publicly reported: Fast loop weighted toward building: a practical agent-building or debugging round, an infrastructure design round, an agent-evaluation round, and a deployed-engineering scenario for customer-facing roles.

Also prepare: the common questions tagged Cognition under Agents and Tool Use, Evaluation and Observability, Coding and Data Structures.

Agents and Tool Use

  • You have eight hours to build a coding agent from scratch. Describe what you build and, more importantly, what you cut.
  • Cognition published an argument against multi-agent systems and later published what actually works. Reconcile those two positions.
  • Your agent spends over half its first turn just finding the relevant code. How do you fix that?

Fine-Tuning, Post-Training and Alignment

  • You are training an agent model with end-to-end RL in your own harness. Walk through the environment and reward design.

AI System Design

  • Design the execution environment for thousands of concurrent cloud coding agents. It must survive the agent waiting forty minutes for CI.
  • Devin runs asynchronously in the cloud; Windsurf's Cascade runs in the editor next to the user. What actually changes between those two products, technically?

Evaluation and Observability

  • How would you evaluate an autonomous software engineering agent? Explain why SWE-bench pass rates mislead.

Safety, Security and Responsible AI

  • An autonomous agent has write access to a customer's repository, CI credentials and network access. What is your threat model?

Applied and Forward-Deployed Scenarios

  • As a Deployed Engineer, you are rolling Devin into a 2,000-engineer organisation. What do the first ninety days look like?

Sierra

Roles this covers: Agent Engineer, Software Engineer, Forward Deployed Engineer, ML Engineer, Product Engineer.

Interview loop, as publicly reported: Recruiter screen → technical screen → a two-hour build session where you can use any AI tools you like → agent-design and evaluation rounds → customer-scenario round → founders/values. Judgment about where to put guardrails is the core signal.

Also prepare: the common questions tagged Sierra under Safety, Security and Responsible AI, AI System Design.

Coding and Data Structures

  • You're handed a small unfamiliar agent codebase. Users report it sometimes confirms an order that was never actually placed. How do you debug it?

RAG and Retrieval

  • The agent answers from a customer's knowledge base, which contains outdated and contradictory articles. How do you prevent confidently wrong answers?

Agents and Tool Use

  • Design a customer-facing agent for an airline that can cancel and rebook flights. How do you keep it from violating fare policy?
  • LLMs are non-deterministic, but a refund over $200 must never be auto-approved. Where's the line between prompting and code?
  • Design the human-handoff path for a customer-service agent. When should it escalate, and what does a good handoff look like?

Evaluation and Observability

  • The space of possible conversations is effectively infinite. How do you evaluate a conversational agent before launch?
  • Your agent passes 92% of eval tasks. Why might that number be misleading, and what would you measure instead?
  • After a foundation-model version upgrade, your production agent's escalation rate doubles overnight. Walk me through your response.

Safety, Security and Responsible AI

Multimodal, Speech and Voice AI

Behavioral and Culture

  • In our build session you get two hours and any AI tools you want. How do you decide what to build and how do you spend the time?
  • Tell me about a time you owned a customer-facing problem end to end.

Harvey

Roles this covers: Software Engineer, ML / Applied AI Engineer, Forward Deployed Engineer, Research Engineer (legal domain).

Interview loop, as publicly reported: Recruiter screen → paired coding round → an architecture presentation round on a legal-workflow system → an evaluation-design round → a customer/partner-scenario round. Domain rigour and grounding matter more than algorithmic puzzles.

Also prepare: the common questions tagged Harvey under RAG and Retrieval, Evaluation and Observability, Coding and Data Structures.

Coding and Data Structures

  • Paired coding: write a chunker for a legal document that never splits a clause and carries enough context that a retrieved chunk is self-contained.

RAG and Retrieval

  • A lawyer asks about a 200-page credit agreement where the operative clause on page 140 depends on a defined term on page 8. How do you build retrieval that gets this right?
  • When would you put a whole contract in the context window instead of retrieving over it? Defend the answer with numbers.

Agents and Tool Use

  • Design an agent that takes a draft NDA and returns a redlined Word document reflecting the firm's playbook, not a chat response.

AI System Design

  • Present the architecture for a workflow reviewing 5,000 contracts against an 18-question diligence checklist, returning a review grid.

Evaluation and Observability

  • A new frontier model is released and scores better on your benchmarks. What happens before it reaches customers?

Safety, Security and Responsible AI

  • Every assertion in a Harvey answer needs to link back to a specific passage. Design the grounding system, and tell me how you would measure the unsupported-claim rate.
  • An agentic research query returns a memo citing a case that was overruled. Where does that get caught?
  • Two partners at the same firm are on opposite sides of a deal. Design the data isolation for that, on top of normal multi-tenancy.

Applied and Forward-Deployed Scenarios

  • Estimate the cost and turnaround of running your diligence workflow over a 5,000-document data room, and tell me which lever you'd pull first.
  • A partner reports that Harvey missed a change-of-control clause in a contract it reviewed. Debug it.

Glean

Roles this covers: Software Engineer (search / ranking), ML Engineer, AI Engineer (agents), Forward Deployed Engineer, Infrastructure Engineer.

Interview loop, as publicly reported: Recruiter screen → coding screen → onsite: a retrieval/ranking design round, a permissions-and-connectors systems round, a practical coding round, an evaluation round, and behavioural.

Also prepare: the common questions tagged Glean under RAG and Retrieval, Fine-Tuning, Post-Training and Alignment.

Coding and Data Structures

  • Merge ranked results from N connector shards into a global top-k, applying a per-user permission filter. Do it efficiently.

Inference, Serving and GPU Performance

  • Walk me through the latency budget of a query: query understanding → retrieval → rerank → LLM answer. Where do you spend and where do you cut?

RAG and Retrieval

Agents and Tool Use

  • Design an agent that takes actions in enterprise tools (file a Jira ticket, draft an email) on a user's behalf. How do you handle permissions and evaluate it?
  • Design agent orchestration across dozens of connected SaaS systems. Where is authorization enforced, and why can it not live in the model?

AI System Design

  • Design a connector framework that syncs content and permissions from 100+ SaaS apps into one index.
  • Glean's ranking leans on a knowledge graph of people, content and activity. How would you build that graph, and how does it improve retrieval beyond embedding similarity?
  • You have dozens of ranking signals and a brand-new tenant with zero interaction data. How do you rank, and how do you improve?

Evaluation and Observability

  • Design the evaluation framework for an enterprise AI assistant when you cannot look at customer data.

Character.AI

Roles this covers: ML Engineer (inference), Research Engineer, Software Engineer (product / safety), Infrastructure Engineer.

Interview loop, as publicly reported: Inference-economics and serving depth, a prompt-construction coding round, a safety-system design round, and product rounds on discovery and engagement.

Also prepare: the common questions tagged Character.AI under LLM Internals and Architecture, Inference, Serving and GPU Performance, Safety, Security and Responsible AI, AI System Design.

Coding and Data Structures

LLM Internals and Architecture

Inference, Serving and GPU Performance

  • Our serving cost is dominated by KV cache, not weights. Get it down by an order of magnitude and tell me what you give up.
  • Dialogues here average around 180 messages. Design the cache that sits between turns.
  • You train natively in int8 rather than doing post-training quantization. Defend that.
  • Estimate what one message costs us to serve, and tell me which lever moves it most.

AI System Design

  • Design discovery and search across millions of user-created characters.

Evaluation and Observability

  • Users complain that characters drift out of persona after a long session. Diagnose it.
  • Engagement metrics and wellbeing metrics disagree. How do you build a system that resolves that?

Safety, Security and Responsible AI

  • Design the safety system for open-ended character chat.
  • When is intervening during decoding better than filtering the finished reply?
  • Design age assurance for a platform where the under-18 experience is fundamentally different.

ElevenLabs

Roles this covers: Research Engineer (speech), ML Engineer, Software Engineer (real-time audio), Forward Deployed Engineer, Infrastructure Engineer.

Interview loop, as publicly reported: A real-time-audio systems round, a speech-model depth round, a practical streaming coding round, a safety round on voice cloning, and a customer-deployment round for FDE roles.

Also prepare: the common questions tagged ElevenLabs under Multimodal, Speech and Voice AI.

Coding and Data Structures

  • Write a service that proxies streaming TTS to a browser and cancels cleanly when the user navigates away.

Inference, Serving and GPU Performance

  • Serving real-time TTS is a different capacity problem from serving a text LLM. Why, and how do you plan capacity?

Safety, Security and Responsible AI

  • Design the safety stack for voice cloning: consent, watermarking and abuse response.

Multimodal, Speech and Voice AI

  • Budget the end-to-end latency for a real-time voice agent. Why is time-to-first-audio a different problem from an LLM's time-to-first-token?
  • Text normalisation is where TTS quality actually dies in production. Walk me through it.
  • Design the dubbing pipeline: an English video becomes Spanish, same speakers, same timing.

Applied and Forward-Deployed Scenarios

  • A hospital group schedules and confirms outpatient appointments by phone, manually, with three staff on a rota. Design what we would build for them.
  • A contact centre wants to replace its IVR with voice agents. Run the engagement.

Abridge

Roles this covers: ML Engineer (ASR / NLP), Research Scientist, Software Engineer (clinical products), Forward Deployed / Implementation Engineer.

Interview loop, as publicly reported: A clinical-ASR depth round, an evaluation-design round (there is rarely a single correct note), a healthcare-integration systems round, and a privacy/compliance round.

Also prepare: the common questions tagged Abridge under RAG and Retrieval, Safety, Security and Responsible AI, Multimodal, Speech and Voice AI.

ML and DL Fundamentals

  • Turn a conversation into billable diagnosis codes. What is the accuracy bar, and how do you build to it?

Inference, Serving and GPU Performance

  • The note should be ready before the clinician leaves the room. Build me the latency budget, and tell me where the money goes.

RAG and Retrieval

  • The patient's chart already lists their medications. How would you use that to improve transcription of drug names, and how would you keep it from backfiring?

Agents and Tool Use

AI System Design

  • Walk me through writing a finished note back into Epic. What goes wrong?

Evaluation and Observability

  • Two good clinicians write different notes for the same visit. So how do you evaluate note quality at all?
  • Edit rate is the obvious measure of clinician trust. What does it hide, and what would you instrument instead?

Safety, Security and Responsible AI

  • A generated note contains a medication the patient never mentioned. Treat that as a safety incident: how do you detect it before a clinician sees it?
  • Clinicians will not sign what they cannot verify. How would you build span-level provenance from every line of the note back to the conversation?
  • PHI is in every audio file, transcript and note you touch. How does that shape the architecture, and what can you send to a third-party model API?

Multimodal, Speech and Voice AI

  • Our audio is a clinic room: two or three speakers, background noise, accents, and a vocabulary full of drug names. How would you build and improve the ASR for that?

Figure AI

Roles this covers: Robotics AI Engineer, Research Engineer (VLA / manipulation), Controls Engineer, Data Engineer (teleoperation), Deployment Engineer.

Interview loop, as publicly reported: Robot-learning depth (VLA, imitation, RL), a data-pipeline design round, a sim-to-real round, a safety-architecture round, and a hands-on numerical/implementation round.

ML and DL Fundamentals

  • Behaviour cloning on teleoperation data has a well-known failure mode. What is it, and what do you do about it on a real humanoid?
  • A whole-body controller trained entirely in simulation has to run on real hardware. What transfers, what does not, and how do you close the gap?
  • Where does reinforcement learning fit on top of imitation learning for manipulation, and what makes the reward the hard part?

Inference, Serving and GPU Performance

  • A colleague wants to move the semantic layer to the cloud so you can use a much bigger model. Walk me through the latency budget.

AI System Design

  • Design the teleoperation data pipeline. Why is data collection the bottleneck in robotics rather than compute?
  • You have 10 hours of demonstrations for a new task and budget for 50 more. How do you decide what to collect, and what return do you expect?

Evaluation and Observability

  • How do you evaluate a manipulation policy when every trial costs robot time and every failure has physical consequences?
  • You ship a policy to 300 robots. It works in the lab and degrades in the field. Debug it.

Safety, Security and Responsible AI

  • Design the safety architecture for a learned whole-body policy operating near people.

Multimodal, Speech and Voice AI

  • What is a vision-language-action model, and how is it different from an LLM with tools?
  • Helix splits into a large slow model and a small fast one. Why not run a single end-to-end network?
  • Explain action chunking. Why predict a sequence of future actions instead of the next one?

Waymo

Roles this covers: Software Engineer (perception, prediction, planning), Research Scientist, ML Infrastructure Engineer, Simulation Engineer, Safety Engineer.

Interview loop, as publicly reported: Coding screen → onsite: two coding rounds (often numerical/vectorised), a perception or planning depth round, an ML-infrastructure or simulation design round, and behavioural. Safety-case reasoning is weighted heavily for senior roles.

Coding and Data Structures

  • In NumPy, compute minADE and minFDE for multi-modal trajectory predictions with variable-length ground truth. No Python loops.

ML and DL Fundamentals

  • Modular perception, prediction and planning, or end-to-end learned driving? Make the case, then tell me what you would actually build.
  • Design the output representation for a behaviour prediction model. What metrics would you gate it on?
  • Where do vision-language models and foundation models genuinely help in an autonomy stack, and where are they a liability?

Inference, Serving and GPU Performance

  • Budget the compute and latency for the onboard stack. What breaks when a model gets bigger?

AI System Design

  • You have hundreds of millions of fleet miles. How do you find and use the rare scenarios that matter?
  • Design a system that finds driving segments similar to a given one across the entire fleet archive.
  • You are opening in a new city. Structure the safety case.

Evaluation and Observability

  • Disengagement rate is a weak safety proxy. How would you actually measure whether the Driver is safe enough to ship?
  • How do you build a simulator you would trust to gate a release?
  • Two days before a release decision, simulation shows a 15% increase in hard-braking events in one scenario cluster. Walk me through what you do.

Multimodal, Speech and Voice AI

  • Why carry lidar, radar and cameras rather than cameras alone? Where would you fuse them?

Forward-Deployed and Enterprise AI

Palantir

Roles this covers: Forward Deployed Engineer, Forward Deployed Software Engineer, Software Engineer, Data Engineer, Deployment Strategist, AIP engineer.

Interview loop, as publicly reported: Recruiter call (~30 min) → technical screen (live coding or HackerRank: a coding challenge, a SQL query and an API task) → onsite of three 60-minute rounds drawn from decomposition, learning, coding, re-engineering and system design → hiring-manager round that revisits weaker areas. Decomposition is the signature round.

Also prepare: the common questions tagged Palantir under RAG and Retrieval, Agents and Tool Use, AI System Design.

Coding and Data Structures

  • Implement a set of shape classes that compute area, then extend them to handle a new shape.
  • Write a SQL query that joins and aggregates across tables to answer a business question.
  • Build a function that fetches paginated data from a REST API, handling page size and total-page logic.
  • Find and fix a double-counting bug in a function that tallies values in a HashMap.
  • Debug a program that models infection spread across a social graph.
  • You inherit an 800-line pipeline script from a previous deployment. It's slow and occasionally produces wrong numbers. The original author is gone. Go.
  • Given exports from three customer systems, each with its own customer records, write code to produce one deduplicated set of entities and explain your design.

RAG and Retrieval

  • Users ask “how many open orders are blocked on a supplier issue?” Plain RAG gets this wrong. Why, and what's the right architecture?

Agents and Tool Use

  • What is an ontology in the Palantir sense, and why put LLM agents on top of one instead of on raw tables and documents?
  • Design an LLM agent that files and updates work orders in a customer's ERP: real writes to a production system. How do you make that safe?

AI System Design

  • Your platform must support multiple LLM providers, including deployments in restricted environments where only some models are available. How do you architect model selection?

Evaluation and Observability

  • How do you evaluate an LLM workflow before and after giving it access to production operations?

Applied and Forward-Deployed Scenarios

  • A freight rail operator loses tens of millions a year to unplanned locomotive downtime. Decompose this into an engineering plan.
  • Design a system to improve traffic in NYC.
  • Design a sync system between two employee record systems.
  • Design a system that lets multiple teams query a shared dataset without exposing raw data.
  • Design an application to catalog and log species while exploring an unfamiliar environment.
  • A customer executive says “the AI keeps getting things wrong” and wants to cancel the pilot. Walk me through your next 48 hours.

Behavioral and Culture

  • Why Palantir, and why this team? Tell me about a time you pushed back on a customer request.
  • Palantir works with defence and intelligence agencies. How do you think about that, and what would you do if asked to build something you're uncomfortable with?

License

   Copyright (C) 2026 Outcome School

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.

关于 About

Your Cheat Sheet For AI Engineering Interviews at Top AI Companies - Questions and Answers.
outcomeschool.com/program/ai-and-machine-learning
aiai-engineeringai-engineering-interviewai-interviewai-interview-questionscompany-wise-prepcompany-wise-questionsinterview-questionsllm-inferencequestions-and-answers

语言 Languages

Markdown100.0%

提交活跃度 Commit Activity

代码提交热力图
过去 52 周的开发活跃度
8
Total Commits
峰值: 8次/周
Less
More

核心贡献者 Contributors