AI engineer jobs: pay, skills and where the jobs are

Updated 5 Nov 2026. Figures from the jobs in our search, not the whole market.

Contents

What an AI engineer does
The skills the ads ask for
Where the jobs are, country by country
Your CV for an AI engineer job
How we read our data

See open AI Engineer jobs Score your CV for these jobs

What an AI engineer does

An AI engineer builds features on top of large language models: an assistant that answers from a company's documents, a tool that reads and sorts incoming mail, an agent that looks up orders. The work is mostly engineering: getting the right information to the model, checking what comes back, keeping cost and waiting time down, and measuring whether it works.

Job titles vary: AI Engineer, GenAI Engineer, LLM Engineer, Machine Learning Engineer. The last one often leans more towards training and running models, but the overlap is large, and we count them as one role.

The skills the ads ask for

We read a sample of the ads every day and count the skills they name. On a computer, hover a skill in the text to see its share; on a phone, the table has the same figures.

Skill	Share of ads	What is expected
Python	58%	Production Python: typed code, async API calls, tests with pytest, packaging and a small web service. Notebooks alone are not enough.
LLM	59%	Call a model API with tool use and structured output. Count tokens and cost; handle rate limits, retries and timeouts.
AI agents	35%	A tool-calling loop with step, time and cost limits, clear tool descriptions, error recovery and a trace of every step.
TypeScript	23%	Enough to ship the AI feature into a web product yourself: a typed API client, streaming to the browser, tests.
CI/CD	37%	Every change runs tests and evals before it ships; a failed gate stops the release.
Git	32%	Branches, reviews and clean history; prompts and eval sets are versioned like code.
SQL	19%	Read and model data well enough to give an agent safe, read-only views of it.
Azure	23%	A very common cloud for AI work in Europe: identity, networking, storage and Azure OpenAI.
AWS	15%	Core services, IAM and networking; Bedrock or SageMaker for models. depends on the provider
Docker	24%	Build small, reproducible images for your services and models; run them locally the way production runs them.
Kubernetes	22%	Deploy and scale a service: deployments, autoscaling, secrets, GPU nodes for model serving.
RAG	28%	Build retrieval end to end: chunking, embeddings, keyword and vector search, reranking, citations, and an eval set to measure it.
LLM evaluation	5%	Eval sets, probes, LLM judges and their limits, regression gates, and cost and latency budgets in CI.
Vector databases	11%	Know one well (pgvector, Qdrant, Weaviate or similar): indexes, metadata filters, re-indexing when the embedding model changes.
LangChain	8%	Useful to know, never the answer on its own. Be able to explain what it does under the hood. depends on the release
LangGraph	6%	A framework for agent graphs. A plus in many ads; the point is the loop and its limits, not the library. depends on the release
MLOps	12%	Ship and run models: versioned data and models, CI/CD, serving, monitoring, rollbacks.
Fine-tuning	4%	Know when it beats prompting and retrieval, how LoRA works in plain words, and how to evaluate the result.
Prompt engineering	15%	Clear instructions, examples, output formats, and prompts kept in version control and tested with an eval set.
FastAPI	2%	A typed Python web API with async handlers, validation and streaming responses.
GDPR	23%	Personal data in prompts, logs and indexes: purpose, minimisation, where it is processed, and deletion.
English	68%	The working language of most AI teams and of all the documentation you will read.

Shares of the AI engineer ads we read in the last 7 days, as of 5 Nov 2026. Our ads, not the whole market.

Three things stand out. First, Python and LLM work lead, as you would expect. Second, plain software skills are asked for a lot: CI/CD, Git, TypeScript, SQL. The job is engineering first. Third, the newer words, agents, RAG, evals, vector databases, appear in fewer ads than the hype suggests, but they are what the interviews test.

The clouds are mostly Azure and AWS. Docker and Kubernetes show up where the team runs its own services.

Where the jobs are, country by country

Each country below has its own figures: which cities have the most jobs in our list, whether the number is growing, how often ads offer remote or hybrid work, and pay ranges from published research, in the local currency.

Germany

15% of the ads offer hybrid work, and 17% are fully remote.

Pay (Base salary)

Entry level: €50,000 to €65,000 a year
Mid-level: €65,000 to €85,000 a year
Senior: €85,000 to €120,000 a year

Total pay, with stock and bonus, usually runs €70,000 to €112,000 a year across levels, and more at big tech companies.

Senior pay at larger employers can reach about €140,000, by one recruiter's estimate.

Munich pays about 10 to 15% more for senior roles and Frankfurt about 5 to 10% (a recruiter's estimate); Baden-Württemberg, Hesse and Bavaria pay the most by state.

Sources: Robert Half 2027 Salary Guide (a projection), Gehalt.de (2026), Levels.fyi (Oct 2026), Payscale (Feb 2026), Indeed (Oct 2026), Optima Search Europe (Apr 2026).

Pay researched on 25 Oct 2026.

Language: many ads in Germany mention German, often for customers, documents and meetings rather than for the engineering itself. Many AI teams work in English. Read the wording: "fließend" means fluent, "gute Kenntnisse" a good working level.

Cloud: many companies reach models through Azure OpenAI, for reasons of contract and where data is kept.

Law (EU): GDPR and the EU AI Act apply. Expect questions on personal data in prompts, logs and search indexes.

Job figures as of 5 Nov 2026.

United States

15% of the ads offer hybrid work, and 40% are fully remote.

99% of the ads that say are full-time.

Pay (Base salary)

Entry level: $100,000 to $150,000 a year
Mid-level: $150,000 to $185,000 a year
Senior: $180,000 to $250,000 a year

Total pay, with stock and bonus, usually runs $170,000 to $370,000 a year across levels, and more at big tech companies.

Big tech pays far more than most employers once stock is counted: the top tenth of total pay is above $500,000.

The San Francisco Bay Area pays the most, then New York; Austin and Seattle also pay well.

Sources: Levels.fyi (Oct 2026), Built In (2026), Robert Half 2027 Salary Guide (a projection), Indeed (Oct 2026), Payscale (Oct 2026), Talent.com (2026).

Pay researched on 25 Oct 2026.

Job figures as of 5 Nov 2026.

United Kingdom

Most AI engineer jobs in our list are in London, Remote job and Bristol.

13% of the ads offer hybrid work, and 37% are fully remote.

Pay (Base salary)

Entry level: £40,000 to £60,000 a year
Mid-level: £65,000 to £85,000 a year
Senior: £95,000 to £130,000 a year

Total pay, with stock and bonus, usually runs £69,000 to £148,000 a year across levels, and more at big tech companies.

London pays about 9% more than the UK as a whole.

Contract work is common: the middle of the advertised day rates is about £500 to £690.

Sources: ITJobsWatch (Oct 2026), Robert Half 2027 Salary Guide (a projection), Levels.fyi (Oct 2026), Payscale (Jul 2026), Indeed (Oct 2026).

Pay researched on 25 Oct 2026.

Job figures as of 5 Nov 2026.

Switzerland

1% of the ads offer hybrid work, and 41% are fully remote.

Pay (Base salary)

Entry level: CHF 85,000 to CHF 105,000 a year
Mid-level: CHF 110,000 to CHF 135,000 a year
Senior: CHF 140,000 to CHF 190,000 a year

Total pay, with stock and bonus, usually runs CHF 120,000 to CHF 229,000 a year across levels, and more at big tech companies.

Zurich pays slightly more than Switzerland as a whole.

Some sources have small samples, so read the senior end with care.

Sources: Robert Half 2027 Salary Guide (a projection), Lohncheck.ch (Apr 2026), Levels.fyi (Oct 2026), Indeed (Oct 2026), Payscale (Sep 2026).

Pay researched on 25 Oct 2026.

Job figures as of 5 Nov 2026.

India

Most AI engineer jobs in our list are in Bengaluru, Hyderabad and hosur road bangalore.

7% of the ads offer hybrid work, and 37% are fully remote.

Job figures as of 5 Nov 2026.

Pay in other countries

Country	Entry level (base)	Mid-level (base)	Senior (base)
Canada	CAD 75,000 to 95,000	CAD 100,000 to 140,000	CAD 130,000 to 180,000
India	INR 4 to 7 lakh	INR 8 to 15 lakh	INR 20 to 30 lakh

These countries have fewer sources, so read them as a rough guide. In India, pay at multinational and big tech companies is far above these ranges. Sources: Levels.fyi (Canada, India), Indeed (Canada, India) and Payscale (Canada, India), 2026. Researched on 25 Oct 2026.

Your CV for an AI engineer job

An AI engineer CV is read by two readers: the hiring software, which looks for the words of the ad, and an engineer, who looks for proof. Write for both.

Use the words the ads use

Name each skill the way the ad names it: "RAG", "LLM", "agents", "evals". Add the long form once if it helps: "retrieval-augmented generation (RAG)".
Put the core skills in a short line near the top: Python, LLM APIs, RAG, evals, your cloud, Docker, CI/CD.
Only list what you can talk about for ten minutes. Every word on the CV is a question in the interview.

Prove each skill

A skill in a list proves nothing. A line under a job proves it: what you built, how big it was, and one result that shows it worked, such as quality, speed, cost or use.

Skill	Weak line	Strong line, with proof
RAG	Experience with RAG.	Built a cited assistant over 12,000 contract pages for 300 staff; 9 in 10 answers fully backed by sources in testing.
Evals	Familiar with LLM evaluation.	Wrote a 300-question eval set and a release gate; it caught 3 regressions before release.
Agents	Built AI agents.	Agent that answers order questions with 3 tools and a step limit; handovers to staff fell by a third.
Python	Python, FastAPI.	Python service (FastAPI) answering 40 requests a second in under 2 seconds.
Cost	Optimised LLM costs.	Cut the cost per answer by almost half with caching and a smaller model for easy questions; quality held.

Never claim what you did not do. A side project is fine: say it is one, link the code, and describe its results the same way.

How we read our data

The jobs: the AI engineer jobs in our job search, matched the same way the search matches them. Our list is not the whole market, so we give shares and rankings, never counts.
The skills: we read a sample of the ads every day and count the skills in them with a fixed word list. We keep only the counts, never the ad text.
Pay: from published salary research, with each source and its date, as a range from entry level to senior. It is not worked out from our ads.
Dates: every figure is worked out again each day, and shows the day it was worked out.

Frequently asked questions

What does an AI engineer do?: Builds features on top of large language models: search over documents, assistants, agents, automatic reading and sorting. Most of the work is engineering and measuring, not training models.
Is an AI engineer the same as a machine learning engineer?: They overlap. AI engineer ads lean towards building products on language models; machine learning engineer ads lean towards training and running models.
Which skills matter most?: Python and work with language model APIs lead, followed by plain software skills such as CI/CD, Git and SQL. Retrieval, agents and evals are what the interviews test.
What should an AI engineer CV show first?: One or two shipped AI features, each with what it did and one result that proves it worked. Then a short skills line in the words of the ads.

More on the AI engineer role

The cvscored teamPublished 25 Oct 2026 · Updated 25 Oct 2026

See open AI Engineer jobs Score your CV for these jobs

Related roles

Jobs from Arbeitnow and the Bundesagentur für Arbeit