Applied AI Jobs in 2026: Mistral and Amazon Are Hiring for Completely Different Things
“Applied AI” sounds like a qualifier, not a job title — the implicit pitch is “we use AI, we just don’t publish papers about it.” The hiring data tells a more specific story. And it’s not one story. It’s two, depending entirely on which company is doing the hiring.
Of the 1,552 open AI/ML roles HireIndex tracks as of June 15, 2026 — across 654 companies — 117 carry the explicit Applied AI tag. That’s 7.5% of the index, ahead of AI Software Engineer and AI Platform, and just behind AI Research. It is not a fringe category. But two companies — Mistral AI and Amazon — post 41% of it, and they are not hiring for the same job.
How many Applied AI jobs are available right now?
117, in this week’s index. Here’s where that sits against the rest of the titles in the index:
| Title | Open Roles | Share of Index |
|---|---|---|
| Machine Learning Engineer | 301 | 19.4% |
| Data Scientist | 202 | 13.0% |
| AI Research | 124 | 8.0% |
| Applied AI | 117 | 7.5% |
| AI Software Engineer | 93 | 6.0% |
| AI Platform | 77 | 5.0% |
| MLOps Engineer | 65 | 4.2% |
| AI Strategy | 43 | 2.8% |
| AI Infrastructure | 31 | 2.0% |
| AI Product Manager | 31 | 2.0% |
| LLM Engineer | 26 | 1.7% |
Applied AI now outranks AI Platform and MLOps combined — a category most candidates aren’t searching for by name, because the word “applied” reads like a description, not a label worth filtering on.
Which companies are hiring the most Applied AI roles?
46 companies post Applied AI roles in the current index. Two of them account for nearly half of it:
| Company | Open Applied AI Roles | Share of Category |
|---|---|---|
| Mistral AI | 28 | 24% |
| Amazon | 20 | 17% |
| OpenAI | 6 | 5% |
| Microsoft | 6 | 5% |
| Cohere | 4 | 3% |
| Snorkel AI | 4 | 3% |
Mistral AI alone outposts every other company in the category by more than 2:1. Strip Mistral and Amazon out and the rest of the market thins fast — a long tail of one- and two-role openings spread across labs, banks, and enterprise software companies. That concentration would matter less if the two leaders were hiring for similar work. They aren’t.
What does Mistral AI’s “Applied AI” job actually involve?
Mistral’s 28 roles are almost entirely “Forward Deployed” titles: Applied AI, Forward Deployed Machine Learning Engineer, Applied AI, Technical Lead, Forward Deployed AI Engineer, Applied AI, Forward Deployed Machine Learning Engineer, Critical and Sovereign Institutions. This is the Palantir playbook, not the research-lab one: embed AI engineers directly inside a client’s environment — government, defense, large enterprise — and build the deployment on-site.
The locations confirm it. Mistral’s Applied AI postings cluster in Paris (15) and remote (5), with single-digit clusters in Singapore (4) and Munich (3), plus individual roles in Abu Dhabi, Montreal, Morocco, and Palo Alto. This isn’t remote-first product engineering. It’s a deployment org built to go wherever a “Sovereign Institution” client needs it.
For candidates, that means the job looks nothing like a typical Machine Learning Engineer posting. Expect client-facing work, travel or relocation, and a mandate to make a model work inside someone else’s infrastructure constraints — not to advance the underlying research.
What does Amazon’s “Applied Scientist” job actually involve?
Amazon’s 20 roles are the opposite model: classic applied research embedded inside existing product and operations teams. Titles include Applied Scientist, Search Ranking, Senior Applied Scientist, Pricing Science, Principal Applied Scientist, Conversational Assistant Modeling & Learning, and Quantum Applied Scientist, AWS Center for Quantum Computing.
These roles sit inside Amazon’s existing business units — Stores, Prime Video, AWS, Ads — and the work is squarely scientific: experimentation, modeling, measurement against an internal roadmap. The locations track Amazon’s main engineering hubs: Seattle, Bellevue, New York, Toronto, plus single roles in London, Cambridge, Palo Alto, and Vancouver. No client deployment, no travel mandate — embedded research against a fixed team, not a fixed client.
Same job title, opposite hiring profile. Apply to “Applied AI” roles based on title alone and you’re applying to two different jobs that happen to share a name.
Where are Applied AI jobs located?
| Location | Open Roles |
|---|---|
| Remote | 54 |
| Paris | 15 |
| London | 12 |
| San Francisco | 9 |
| Toronto | 8 |
| Singapore | 5 |
| Seattle | 5 |
| Munich | 4 |
| New York | 4 |
Remote leads, but only at 46% of the category — well below the 63% remote share across the full HireIndex index. Applied AI is structurally less remote-friendly than the average AI/ML role, which tracks with the two hiring models above: forward-deployed engineering and embedded applied science both tend to require physical proximity to a client site or a specific team.
Paris leads non-remote markets almost entirely on the strength of Mistral’s headquarters presence. London splits between Amazon and a handful of enterprise AI teams.
Is Applied AI accessible to early-career candidates?
Barely. The seniority breakdown across the 117 roles:
| Level | Count | Share |
|---|---|---|
| Mid-level | 84 | 71.8% |
| Senior-heavy (Staff / Principal) | 17 | 14.5% |
| Senior | 14 | 12.0% |
| Leadership | 1 | 0.9% |
| Junior | 1 | 0.9% |
One junior role out of 117. Senior+ share sits at 27.4% — lower than LLM Engineer (38%), but the practical floor is “mid-level,” not “entry.” Both hiring models explain why: forward-deployed work requires enough production experience to be trusted alone inside a client’s infrastructure, and embedded applied science requires enough research grounding to design and defend an experiment without close supervision.
If you’re early-career and targeting this category, Machine Learning Engineer is still the more realistic entry point — it carries a real junior slice and builds the production and modeling experience that Applied AI roles assume on day one.
What this means for your job search
Read past the title. “Applied AI” tells you almost nothing on its own. Check the company before the job description — a Mistral posting and an Amazon posting with similar-sounding titles lead to fundamentally different work and interview processes.
Decide which model you actually want. Forward-deployed work means client exposure, travel, and ambiguity. Embedded applied science means a fixed team and deep ownership of a narrow problem. They reward different people.
Don’t expect remote by default. At 46% remote, Applied AI is one of the more location-bound categories in the current index. If flexibility is the priority, browse Remote roles directly rather than assuming an Applied AI search will surface them.
Source matters here. More than half of Applied AI postings surface through aggregator feeds rather than a direct ATS — harder to find by going straight to a company’s careers page.
Browse current Applied AI roles alongside Machine Learning Engineer and AI Research openings. Data updated every Monday from direct ATS and aggregator sources.
HireIndex tracks 1,552 open AI/ML roles across 654 companies as of June 15, 2026.