New York's AI companies are hiring more salespeople than engineers

This is the first issue of a weekly read on who is actually hiring in New York AI.

Every number below comes from the companies' own job boards, read fresh each morning. Not an aggregator, not a survey, not a press release. When a role is filled it disappears from here, because it disappeared from theirs.

Here is where the market stands.

1,550 open roles

Across 48 New York AI companies. 734 of those sit in the city itself, 265 are remote-eligible, and the rest are at New York companies hiring elsewhere.

And it barely moved. Comparing only the 46 companies we tracked every day this week — the like-for-like set — open roles went from 1,507 to 1,513. Six roles. Four tenths of one percent.

The raw totals show +43, but 37 of those arrived because two more companies entered our coverage, not because anyone posted a job. We will always give you the like-for-like number, because the other one flatters us and tells you nothing.

Twelve of those 46 companies moved at all this week. The median move was one role.

A flat week is not a boring week — it is the base rate you need before a busy one means anything.

The finding: they are hiring more salespeople than engineers

530 open go-to-market roles against 474 in engineering.

Sit with that for a second. These are companies whose entire product is artificial intelligence, and they are recruiting more people to sell and deploy it than to build it.

The prevailing story about AI hiring is a talent war for researchers and infrastructure engineers. The boards say the constraint has already moved somewhere else. The hard part is no longer building the thing — it is getting it into an enterprise and keeping it there.

If you are a commercial operator who can hold a technical conversation, this is the least contested part of the New York AI market right now, and it is not close.

You can see the same finding wearing an engineering title: forward deployed engineer is now 40 open roles, larger than AI engineer (10), solutions architect (8) and developer relations (3) put together. That job barely existed two years ago.

Who is hiring hardest

Two companies stand out for a reason the raw count hides. Ramp and EliseAI are hiring in New York, not merely from it — 75% and 87% of their open roles are in the city. ElevenLabs has the biggest board on this list and 4% of it is local.

If you want to work in an office in New York rather than remote for a New York company, that column is the one to read.

What it pays

From salary ranges posted on New York listings. The city's pay transparency law requires a good-faith range on roles performed here, and 85% of NYC postings carry one — which is the only reason anyone can publish this honestly.

By level: staff and principal engineers post a $242,500 median against $190,000 for senior. Base salary only — equity excluded, on-target earnings excluded and counted separately, because bundling commission into a base figure is how salary data gets useless.

More on what these roles pay across levels in the NYC AI engineer salary guide.

The entry-level problem

54 entry-level roles. 189 at staff and principal.

Senior demand is running three and a half times ahead of junior across the entire index.

If you are advising someone trying to break into AI in New York, that single ratio is the most useful thing you can tell them — and it argues for arriving laterally from adjacent engineering work rather than applying straight at a junior AI title. The doors that are open are not the ones most people are knocking on. We wrote up the honest routes in how to get into AI.

What they are building with

Tools named in job descriptions, counted by how many of the 48 companies mention each:

Anthropic appears at 25 New York AI companies against OpenAI's 19.

One week is not a trend and we are not going to pretend otherwise. But it is the kind of thing that is invisible in press releases and obvious in job descriptions, and we will report it each week either way.

Time to fill

Our first closures landed this week. Mirage filled two roles in 48 hours.

That is a sample of two. A median time-to-fill needs a quarter of closures before it means anything, and we will not publish one until it does. We are starting the clock in public so you can watch it fill.

---

Next week: the first read on whether the go-to-market lead is widening or was a blip.