How to get a job at xAI

The single most useful fact about working at xAI is not in any careers page. It is that 49 of their 254 open roles are in Memphis, Tennessee, more than in Palo Alto (69). If you pictured a Bay Area research lab, you have the wrong company in mind.

Memphis is where Colossus lives, xAI's supercomputer cluster. Their largest single team by open headcount is Data Center (124 roles), which is more than any research or product group. This is a company currently hiring more people to build and run physical infrastructure than to do anything else.

What that means for you

If your background is electrical, mechanical, HVAC, power systems, network installation, site operations or industrial construction, xAI is one of the most accessible frontier AI employers in the world right now, and almost nobody in those trades is looking at AI companies. The competition for those roles is not other machine learning researchers.

If you are a researcher, the picture is inverted. Their research and engineering headcount is real but modest next to the infrastructure build, so expect a narrow funnel and a high bar.

The other unusual team

Human Data (28roles) is xAI's data annotation and evaluation organisation, the people who generate and grade the examples models learn from. These roles, often titled AI Tutor, hire for subject expertise rather than for engineering: the point is that you know a domain well enough to judge whether a model's answer is any good.

They are frequently remote, which is most of why 28of xAI's roles are listed that way. If you have deep expertise in medicine, law, finance or a science and no intention of writing code, this is a genuine route in.

How to read their listings

xAI publishes less about its process than its peers. There is no public interview guide, no stated values framework and little advice to applicants, so treat anyone claiming to describe their loop in detail with caution.

What their listings do consistently signal is a preference for demonstrated building over credentials, and a tolerance for speed over process. The company's public reputation is for moving quickly with small teams, and the job descriptions read that way.

The practical version

• Check Memphis and the other non-Bay-Area sites before assuming there is nothing near you.
• If you are a domain expert rather than an engineer, look at the Human Data roles.
• Infrastructure and datacenter experience is worth more here than at any other lab on this list.
• Expect less structure and less published guidance than at Anthropic or DeepMind.

Team and location counts are read from xAI's live job board and refresh daily. xAI publishes little about its hiring process, so this guide deliberately makes fewer claims about interviews than the others and sticks to what the listings themselves show. See their open roles.