Google DeepMind interview, decoded.

Updated · Sources at the end

Google DeepMind runs a role-specific loop that goes deeper on machine learning fundamentals and past research than a standard Google interview: expect strong coding, real ML theory, and a discussion of your own work. Google's AI research organization, building frontier models and pursuing AGI, it hires research scientists, research engineers, and software engineers.

Free · 2 minutes · no account

The questions for your exact Google DeepMind role and level.

01

Who Google DeepMind hires

The main roles are research scientist, research engineer, and software engineer, and DeepMind favours candidates with strong ML fundamentals, solid coding, and a track record of research or hands-on ML projects, with research scientists leaning toward PhD-level publication experience.

Wondering whether your resume reads like someone Google DeepMind hires? Get a free resume review.

There are three main tracks, each with its own tilt on the same loop. All of them expect strong ML fundamentals and solid coding, and the difference is where the depth rounds point.

  • Track · RS

    Research scientist

    PhD-level publication record expected. The loop leans hardest on your own papers and your research thinking.

  • Track · RE

    Research engineer

    Hands-on ML projects and engineering craft. Paper discussion plus serious coding and experimentation depth.

  • Track · SWE

    Software engineer

    Closest to a standard Google loop, with strong unaided coding and ML fluency as the differentiator.

02

What is the Google DeepMind interview process?

The loop runs through Google's hiring machinery, so expect recruiter and technical screens followed by a five to seven round panel and a hiring committee, with the exact rounds tuned to whether you are on the research, research-engineering, or software track.

6 roundsRecruiter and hiring manager screen · Coding rounds · ML depth and theory · Research or past-work discussion · Team lead interview · Behavioural and cultureSee each roundHide
01
Recruiter and hiring manager screen
Phone or video call

Background, motivation, and which team or track fits, including why DeepMind and interest in its mission.

02
Coding rounds
One to two live coding sessions

Data structures and algorithms at medium to hard difficulty, solved unaided, with your approach explained before you write code.

03
ML depth and theory
Technical interview

Deep learning fundamentals plus the underlying math in linear algebra, probability, and optimization, sometimes with ML system design.

04
Research or past-work discussion
Presentation and deep dive

Present a paper or your own projects and defend the methods, results, and next steps against probing questions.

RS & RE tracks only
05
Team lead interview
Conversational technical chat

Your experience with ML experimentation and modelling, open-ended ML problems, and how you would fit the team.

06
Behavioural and culture
People and culture interview

Mission alignment, collaboration, handling disagreement, and end-to-end ownership of projects.

03

What Google DeepMind screens for

Across rounds, DeepMind is reading for a consistent signal on a few core traits.

  • Depth in machine learning fundamentals and the math behind them
  • Research thinking and the ability to frame open-ended problems
  • Strong, unaided coding and engineering craft
  • Collaboration across research and engineering teams
  • Genuine alignment with the mission of building AI responsibly
04

Google DeepMind interview questions

Candidate-reported themes

The questions below reflect the kinds of themes candidates report across DeepMind loops.

Rehearse these out loud for Google DeepMind. Free to try.

See what's included →

Behavioural & motivation

  • Why DeepMind, and which of our research directions excites you most?
    Cal

    I'm Cal, your AI coach. Say whatever comes to mind: I'll put your answer in order, show what's missing, then score it.

    Scored on Specificity · Connection · Research
    • Your answer is scored on
    • SpecificityNames at least two concrete things about this company or role that could not be copy-pasted to a competitor.
    • ConnectionTies those specifics to their own goals or experience, so it reads as fit rather than flattery.
    • ResearchShows they actually looked: a product, a technical choice, a public post. Generic praise for culture or mission does not count.
    See how a strong answer sounds

    Example · not your answer

    “You're the only team I've found that publishes incident reviews with the real timeline in them, and I've read four now. I've worked somewhere that treated postmortems as a ritual you get through, and I'd rather not do that again.”

    Strong. One row still open.

    • SpecificityOne, not twoThe incident reviews are the only specific here. The rest could be said to any company.
    • ConnectionJoined upTies the observation to how they want to work, not just to admiration.
    • ResearchLooked properlyRead four of them. That is looking, not skimming an About page.

    That comes back in about ten seconds, with what worked, what to fix, and what the question was really after.

    We never store your recording or your answer. How this is handled ↗

  • Tell me about a project you owned end to end and what you would do differently.
  • Describe a time you disagreed with a collaborator on a technical decision and how you resolved it.
  • How do you handle a sudden shift in research priorities mid-project?

How to structure your answer out loud

Technical

  • Design a Trie for prefix matching, with insert, search and a startsWith query. Explain your approach before any code.Algorithmic coding: A Trie for prefix matching, a snapshot array, medium-to-hard data structuresListening for: Sound structure · Complexity stated · Explains before coding
  • Derive backpropagation for a two-layer network and explain how the bias-variance trade-off shows up when you train it.ML theory: Deriving backpropagation, the bias-variance trade-off, comparing loss functionsListening for: Correct derivation · Bias and variance · Clear and structured
  • Explain the singular value decomposition, how it relates to eigenvalues, and where you would use it in machine learning.Math foundations: Eigenvalues and SVD, Bayes rule, multivariate optimizationListening for: Correct definition · Link to eigenvalues · Real ML uses
  • Present a paper or project of yours, then defend its methodology, its weaknesses and how you would extend it.Your own work: Defending your paper or project's methodology, weaknesses and extensionsListening for: Clear methodology · Weaknesses owned · Next steps
Your Google DeepMind interview

Don’t walk in hoping. Walk in rehearsed.

Calibrd predicts the questions Google DeepMind is likely to ask for your role, interviews you out loud, and tells you what a hiring manager would hear.

Free · talk any answer through

Ready to go all in? for 90 days · or Unlimited graded answers and 45 mock interviews · no subscription ·
Built by Xiaochen Guo, 500+ interviews on the hiring side

How the questions are predicted

05

Pay

L4~$300K total pay, US
L5$400–600K total pay, US
L6Higher again, senior staff and up

Pay tracks Google's levels. London packages typically run well below Bay Area figures, and Calibrd benchmarks the pay on any posting you scan against its level and location.

Applying to Google DeepMind? Paste the posting and see whether they’d shortlist you →

Already receiving LinkedIn alerts for Google DeepMind? Forward them to your own Calibrd address. Calibrd reads each job against your resume and keeps only the ones you could be shortlisted for, so you apply where your odds are best. Set it up.

06

How to prepare for a Google DeepMind interview

  1. Rehearse a clear ten-minute walkthrough of one of your papers or projects, and be ready to defend the methods, results, and limitations.
  2. Drill medium to hard algorithm problems and plan to solve them without AI assistants, since technical rounds are generally unaided.
  3. Refresh the core math and ML theory: linear algebra, probability, optimization, backprop, and common architectures and loss functions.
  4. Read the recent work of your interviewers and DeepMind's teams so you can connect your background to their research.

This guide covers Google DeepMind's engineering and research hiring. For management and leadership roles the loop is similar but the bar shifts to people, delivery and strategy, so pair it with the leadership interview prep hub. The bar for your exact role comes from the role-by-role guides, and the prep that actually transfers is spoken, so run a mock interview before the real one.

Also interviewing at Mistral AI, Cohere or xAI? Their guides follow the same format, from the first screen to the offer.

Knowing the questions isn’t the same as answering them out loud. Run a Google DeepMind mock: spoken answers, coached on the spot. Your first mock is free.

Run a Google DeepMind mock →
07

FAQ & sources

What is Google DeepMind's interview process?

The loop runs through Google's hiring machinery, so expect recruiter and technical screens followed by a five to seven round panel and a hiring committee, with the exact rounds tuned to whether you are on the research, research-engineering, or software track. The stages, in order: Recruiter and hiring manager screen; Coding rounds; ML depth and theory; Research or past-work discussion; Team lead interview; Behavioural and culture.

What does Google DeepMind look for in candidates?

Across rounds, DeepMind is reading for a consistent signal on a few core traits. Depth in machine learning fundamentals and the math behind them; Research thinking and the ability to frame open-ended problems; Strong, unaided coding and engineering craft; Collaboration across research and engineering teams.

What questions does Google DeepMind ask in interviews?

The questions below reflect the kinds of themes candidates report across DeepMind loops. Technical themes: Algorithmic coding; ML theory; Math foundations; Your own work.

How much does Google DeepMind pay?

L4: ~$300K total pay, US; L5: $400–600K total pay, US; L6: Higher again, senior staff and up. Pay tracks Google's levels.

How do I get a job at Google DeepMind?

The main roles are research scientist, research engineer, and software engineer, and DeepMind favours candidates with strong ML fundamentals, solid coding, and a track record of research or hands-on ML projects, with research scientists leaning toward PhD-level publication experience. Hiring runs in 6 stages, starting with Recruiter and hiring manager screen.

How do I get job alerts for Google DeepMind jobs?

Already receiving LinkedIn alerts for Google DeepMind? Forward them to your own address in Calibrd's Job search. Calibrd reads only what arrives there, nothing else in your mail. Each job in them is checked against your CV in your next search, and you see only the ones you could be shortlisted for. Not receiving any yet? Set one up for Google DeepMind on LinkedIn first.

How do I prepare for a Google DeepMind interview?

Rehearse a clear ten-minute walkthrough of one of your papers or projects, and be ready to defend the methods, results, and limitations. Drill medium to hard algorithm problems and plan to solve them without AI assistants, since technical rounds are generally unaided.

Prep for a real Google DeepMind role

Practise your Google DeepMind interview, out loud.

Paste a real Google DeepMind posting and Calibrd predicts the questions for that role and level, benchmarks the pay, and flags the gaps an interviewer will probe in your resume — then listens to your spoken answers and coaches them. Your first mock is free.

Free to start · No card · Encrypted at rest, never used to train AI, remove anytime

Two weeks out?

Email me the Google DeepMind prep checklist

The round that decides it, the questions they ask, and what to do before, in one email. Nothing else unless you sign up.

Google DeepMind Interview: Process, Paper Defence — Calibrd