AI Research Scientist vs. ML Engineer: Who's Actually Hiring Which

Two titles dominate AI job listings right now: Machine Learning Engineer and Research Scientist. Candidates confuse them constantly. So do job boards.

They’re not the same job. The companies hiring for each are largely different. The interview process is different. The career trajectory is different.

Here’s what the actual hiring data shows.

How many ML Engineer vs Research Scientist jobs are available?

In our current index of 448 open AI/ML roles across 89 companies:

ML Engineer titles: 13.8% of all indexed roles (~62 roles)
Research Scientist titles: 6.7% (~30 roles)
AI Engineer titles: 10.9% (~49 roles) — and climbing fast
Research Engineer: 3.1% (~14 roles)

ML Engineer roles outnumber Research Scientist roles roughly 2:1. If you’re only applying to one, you’re competing in whichever pool has more people looking the wrong direction.

What is the difference between ML Engineer and Research Scientist?

ML Engineer builds systems. The job is getting models into production — training infrastructure, inference pipelines, latency optimization, feature stores, eval frameworks. The output is something that ships and runs. Most ML engineers work on products, not papers.

Research Scientist advances the state of the art. The job is finding things that don’t work yet and making them work. The output is usually a paper, a new technique, or a demonstration that something is possible before the engineering team takes it from there. Most research scientists have PhDs. Many came from academia.

AI Engineer is the newer title — and based on the 65% relative growth we’ve seen in three weeks, it’s becoming the dominant one. In practice it sits between the two above: takes frontier models (GPT, Claude, Gemini, Llama) and builds applications and products on top of them. Heavy emphasis on prompt engineering, retrieval systems, and tool use. A PhD is rarely expected.

Research Engineer is the rarest. It’s the hybrid: rigorous enough to run experiments, systems-minded enough to implement them cleanly. Usually found inside dedicated research labs (DeepMind, MSR, FAIR) rather than product teams.

Which companies hire ML Engineers vs Research Scientists?

Research Scientists are concentrated at a handful of organizations. If you’re looking for this role, the search space is narrow:

OpenAI, Anthropic, DeepMind, Google Brain (now Google DeepMind), Meta FAIR, Microsoft Research — these are the pure-play research labs. Most positions here require a PhD and a publication record.
Databricks and Scale AI both run applied research teams. The bar is high, the volume is low.

In our current index, three companies account for more than half of all Research Scientist openings. This is not a distributed market. It’s oligarchic.

ML Engineers are everywhere. This is the role that product companies, startups, and enterprises all hire for in volume:

Scale AI and Databricks dominate, as they do in most AI categories.
Mid-sized companies like PointClickCare, Coreweave, and Snorkel AI have multiple ML engineering seats open.
Series A/B startups on Ashby are disproportionately ML Engineering-heavy — they need people who ship, not people who publish.

If you have an ML background and you’re not sure whether to position yourself as a researcher or an engineer, the hiring volume answers the question: engineer roles are twice as common and distributed across far more companies.

What does the interview process look like for each role?

The fastest way to tell the roles apart before you apply: look at the interview process the company describes and what they ask for in the application.

Research Scientist listings typically say: research statement, publication list, references from faculty or senior researchers. If you don’t have those, the role is not for you regardless of your technical skills.

ML Engineer listings ask for: system design, coding, ML fundamentals, take-home projects or case studies. PhD preferred but usually not required. Production experience weighted heavily.

AI Engineer listings lean even more applied: portfolio of shipped AI products, familiarity with specific LLM APIs, sometimes a live coding assessment using OpenAI or Anthropic tools.

Where can I find ML Engineer and Research Scientist job listings?

Browse the current data by skill:

The index updates every Monday with fresh listings pulled directly from Greenhouse, Lever, Ashby, and SmartRecruiters — before they’re syndicated to the boards where everyone else is searching.

Data from HireIndex’s live pipeline as of early May 2026. 448 open roles, 89 companies, updated weekly.