How to get a job at Google DeepMind

Start with a caveat that matters more than anything else here: the 0roles we track for Google DeepMind are only part of their hiring. Most Google DeepMind positions are posted through Google's own careers site rather than the board we read. If you search only here, you will miss the majority of what is open.

We show what their public board exposes and link straight to it. For a complete picture, search Google Careers for DeepMind as well. No aggregator, including this one, sees everything.

They publish their interview structure

Google DeepMind sets out four stages. First, initial interviews: a thirty-minute recruiter call about your background and the role, sometimes followed by a hiring manager conversation about specific capabilities. Second, skills interviews: two to three calls assessing the competencies the role needs, where you meet people you would work with.

Third, final interviews with team leads and leadership, which they frame explicitly as “a mutual assessment.” Fourth, a decision and offer, with the hiring team reviewing candidates against their criteria.

The framing of the final stage as mutual is worth taking at face value rather than as politeness. Arriving with real questions about the team's direction is expected behaviour there, not a bonus.

Research scientist and research engineer are different jobs

DeepMind draws this line more clearly than most. Research Scientists “formulate novel hypotheses and algorithmic paradigms” and typically hold PhDs. Research Engineers bridge “theory and implementation” through experimental work.

If you are a strong engineer without a doctorate, the research engineer track is the one written for you, and applying to the scientist track instead is a common and avoidable mistake. Their software engineering, product management, program management and operations roles sit outside the research ladder entirely.

Where they are

Google DeepMind operates from London, the Bay Area, Bangalore, Cambridge in Massachusetts, Montreal, New York, Paris, Tokyo, Toronto and Zurich. London is the historical home and remains substantial, which makes them one of the few labs where Europe is a first-class location rather than a satellite.

They state no explicit remote policy, which in practice means assume you are expected in an office unless a listing says otherwise.

If you are early in your career

They run a Student Researcher Program placing people alongside their research staff. For anyone still studying, that is a materially easier door than the open roles, and it is advertised separately from them.

The practical version

• Search Google Careers as well as any aggregator, including this one.
• Pick the research engineer track, not research scientist, if you do not have a PhD.
• Bring genuine questions to the final round, because they have called it a mutual assessment.
• London is a real centre here, not a branch office.
• If you are a student, the Student Researcher Program is the easier route.

Process, role definitions and locations are from Google DeepMind's careers pages. Counts are read from their public job board and refresh daily, and as noted above represent only part of their open hiring.