All reports
AI Engineer · employed per open role

AI Engineer: employed per open role

IBM leads the multiple at 44.6× (759 employed / 17 open); Capital One posts the most open roles (366). As of 2026-11-14 PT.

As-of 2026-11-14 PT Published by Metix AI Collection series Titles measured: AI Engineer / Senior AI Engineer
IBM NVIDIA Databricks Capital One Microsoft 44.6× · 759 / 17 8.5× · 136 / 16 5.9× · 59 / 10 0.25× · 90 / 366 241 employed · open roles not captured AI Engineer employed per open role As of 2026-11-14 PT IBM NVIDIA Databricks Capital One Microsoft 44.6× · 759 / 17 8.5× · 136 / 16 5.9× · 59 / 10 0.25× · 90 / 366 241 employed · open roles not captured AI Engineer employed per open role As of 2026-11-14 PT AI Engineer employed per open role. IBM 44.6× (759 employed / 17 open); NVIDIA 8.5× (136 employed / 16 open); Databricks 5.9× (59 employed / 10 open); Capital One 0.25× (90 employed / 366 open); Microsoft 241 employed, open roles not captured. As of 2026-11-14 PT. Source: Metix AI.
AI Engineer employed per open role. Microsoft 241 employed, open roles not captured. Denominators: IBM 17; NVIDIA 16; Databricks 10; Capital One 366. As of 2026-11-14 PT. Source: Metix AI.
Data & notes
IBM 44.6× (759 employed / 17 open); NVIDIA 8.5× (136 employed / 16 open); Databricks 5.9× (59 employed / 10 open); Capital One 0.25× (90 employed / 366 open);
CompanyEmployedOpenEmployed per open role
IBM7591744.6×
NVIDIA136168.5×
Databricks59105.9×
Capital One903660.25×
Microsoft2410not captured
Named list

Read the multiple first

Take the named list for the employers shown here. Keep the title family this page counted.

Key takeaway · How to cite
Where the open roles sit

Capital One posts the most open AI Engineer roles: 366

Start with where demand sits, as of 2026-11-14 PT.

0% 20% 40% 60% 80% 100% 89.5% Capital One 366 (89.5%) IBM 17 (4.2%) NVIDIA 16 (3.9%) Databricks 10 (2.4%) Open AI Engineer roles by host As of 2026-11-14 PT 0% 30% 60% 90% 89.5% Capital One 366 (89.5%) IBM 17 (4.2%) NVIDIA 16 (3.9%) Databricks 10 (2.4%) Open AI Engineer roles by host As of 2026-11-14 PT Captured open AI Engineer roles by host: Capital One 366, IBM 17, NVIDIA 16, Databricks 10. Hosts with uncaptured open roles are left out. As of 2026-11-14 PT. Source: Metix AI.
Captured open AI Engineer roles by host. Hosts with uncaptured open roles are left out. As of 2026-11-14 PT. Source: Metix AI.
Data & notes
Capital One 366, IBM 17, NVIDIA 16, Databricks 10.
NameValueShare
Capital One36689.5%
IBM174.2%
NVIDIA163.9%
Databricks102.4%
Employed per open role

Which hosts rank highest on employed per open role?

The ratio chart above ranks every host.

HostEmployedOpen
IBM75917
NVIDIA13616
Databricks5910
Capital One90366
Microsoft241not captured
Named list

Get the named AI Engineer list behind this table

Point the list request at the employers and wires already counted here. Name the same employers and experience cut this page measured.

Bench depth

Who holds the deepest bench? IBM at 759

Employed headcount ranked on its own; IBM is largest.

IBM Microsoft NVIDIA Capital One Databricks 0 200 400 600 800 759 241 136 90 59 AI Engineer employed As of 2026-11-14 PT IBM Microsoft NVIDIA Capital One Databricks 0 300 600 900 759 241 136 90 59 AI Engineer employed As of 2026-11-14 PT AI Engineer employed depth, largest first. IBM employed 759; Microsoft employed 241; NVIDIA employed 136; Capital One employed 90; Databricks employed 59. As of 2026-11-14 PT. Source: Metix AI.
AI Engineer employed depth, largest first. As of 2026-11-14 PT. Source: Metix AI.
Data & notes
IBM employed 759; Microsoft employed 241; NVIDIA employed 136; Capital One employed 90; Databricks employed 59.
NameValue
IBM759
Microsoft241
NVIDIA136
Capital One90
Databricks59
Employed vs open

Absolute employed beside opens: ratio lead IBM, bench lead IBM

The two rankings can diverge: bench volume and door density are different decisions.

0 200 400 600 800 IBM NVIDIA Databricks Capital One Microsoft 759 136 59 90 241 17 16 10 366 0 Employed Open roles AI Engineer: employed vs open by host As of 2026-11-14 PT 0 300 600 900 IBM NVIDIA Databric ks Capital One Microsof t 759 136 59 90 241 17 16 10 366 0 Employed Open roles AI Engineer: employed vs open by host As of 2026-11-14 PT Absolute employed beside captured open roles. Ratio lead IBM (44.6×); bench lead IBM (759); open lead Capital One (366). As of 2026-11-14 PT. Source: Metix AI.
Absolute employed beside captured open roles. As of 2026-11-14 PT. Source: Metix AI.
Data & notes
Ratio lead IBM (44.6×); bench lead IBM (759); open lead Capital One (366).
CategoryEmployedOpen roles
IBM75917
NVIDIA13616
Databricks5910
Capital One90366
Microsoft2410
How to use it

What it means for AI Engineer employed vs open: what to do this week / how to use it

How to read it

How to read hosts without captured open roles

Open roles were not captured at Microsoft. Those hosts are listed in grey, with the employed count and no bar, and they are excluded from employed per open role.

There is no five-host peer-pack ladder here: the comparison is employed per open role on one seat.

As of 2026-11-14 PT · Source: Metix AI

Notes & method

Methodology

The employed cell is current AI Engineer headcount on that company's homepage record at 2026-11-14 PT. The open cell counts postings still open for this title family. An open cell of 0 is uncaptured, not a claim of no demand. Source: Metix AI.

Aggregate report · no personal information shown · provided by Metix AI
FAQ

FAQ

Does an uncaptured AI Engineer host get a bar?
If open roles were not captured, that host is shown without a ratio bar. The cell 0 is not captured, rather than a claim of no demand.
Is the AI Engineer page a peer-pack ladder?
It is an employed-per-open reading for AI Engineer, not a five-company ladder. The ratio skips hosts that were not captured.
Who leads the AI Engineer multiple?
The top of the ratio list is IBM 44.6× (759 employed / 17 open). That multiple is dated 2026-11-14 PT.
Where do the AI Engineer figures come from?
The employed count and the open count are as of 2026-11-14 PT. The source is Metix AI.

Request the AI Engineer open-req list

Ask for people on the measured wires, not a broader uncut list. The request should name the same employers.

Optional: try Mira
中文版