AI Engineer, ML Engineer, Applied AI, LLM: Job Titles to Watch
AI hiring comes with a pile of overlapping job titles. The same work can be called AI Engineer at one company, Applied AI Engineer at another and "Software Engineer, Machine Learning" at a third, while two jobs both called ML Engineer can be completely different. This guide groups the common titles by the work behind them and shows how to tell them apart from the posting itself.
Why the titles are messy
AI job titles moved faster than company job ladders. Many teams started building features on top of large language models before they had a name for the people doing it, so they reused existing titles or invented new ones. The title tells you the area; only the description tells you the job.
Four families of AI titles
1. Building products with models
Common titles: AI Engineer, Applied AI Engineer, LLM Engineer, AI Product Engineer, "Software Engineer, AI".
The work is mostly software engineering: connecting models, often through an API, to a product; building retrieval (RAG) pipelines, agents and tool use; writing evaluations; managing latency and cost. You usually do not train models from scratch. A strong backend or full-stack engineer who has shipped LLM features is often exactly who these postings want.
2. Training and improving models
Common titles: Machine Learning Engineer, ML Engineer, Applied Scientist, Research Engineer, Research Scientist.
Here you work closer to the model: training, fine-tuning, feature engineering, ranking and recommendation systems, experiments. Postings mention PyTorch or similar frameworks, datasets and metrics. "Research" titles tend to lean toward new methods and papers; "Applied" titles toward getting models into products. An ML Engineer at a company with a recommendation system may have nothing to do with LLMs at all.
3. ML infrastructure
Common titles: MLOps Engineer, ML Platform Engineer, ML Infrastructure Engineer, Inference Engineer.
This is platform work for models: training pipelines, model serving, GPU scheduling, monitoring. If you come from DevOps, Platform or SRE, these titles are the natural bridge into AI.
4. Data science with an AI angle
Common titles: Data Scientist, Applied Scientist, Data Scientist (Machine Learning).
More analysis and modelling than production engineering: statistics, experiments, forecasting, and increasingly evaluating model outputs. Where this family ends and family 2 begins differs from company to company.
Adjacent roles
- Solutions Engineer or Forward Deployed Engineer on an AI product: helping customers build with the company's models or tools. Technical and customer-facing.
- AI Product Manager: product roles for AI features, usually listed under Product Manager.
- Developer Relations for AI products: docs, demos, community.
Quick reference
| Title family | Main work | Words to look for |
|---|---|---|
| AI / Applied AI / LLM Engineer | Product features on top of models | RAG, agents, evals, API, TypeScript, Python |
| ML Engineer / Applied Scientist | Training and tuning models | PyTorch, fine-tuning, ranking, experiments |
| MLOps / ML Platform | Infrastructure for models | serving, pipelines, GPUs, Kubernetes |
| Data Scientist | Analysis and modelling | statistics, A/B tests, forecasting |
How to read an AI posting
Because titles are unreliable, read the description with these questions in mind:
- Does the team train models, or use them? Look for "fine-tune", "pre-train" and "training data" versus "integrate", "API" and "product features".
- What is the main language? Python only suggests model work; TypeScript or Go alongside Python often means product engineering.
- What do they measure? Model accuracy and offline metrics point to ML; latency, cost and quality for users point to AI engineering.
- Who will you work with? Researchers, or product managers and designers?
- What does "AI experience" mean to them? Sometimes a PhD, sometimes one shipped LLM feature. The requirements list usually makes this clear.
Titles to track
To avoid missing jobs because of naming, track:
- AI Engineer, Applied AI Engineer, LLM Engineer
- Machine Learning Engineer, ML Engineer, Applied Scientist
- MLOps, ML Platform, ML Infrastructure
- General Backend and Software Engineer postings. Product teams often hire for AI work under a plain engineering title, with the LLM part only in the description.
Watch the location line too, since AI roles are often tied to particular offices or regions. In our September 2026 snapshot of the boards we check, there were 9 AI Engineer and 14 Machine Learning postings open to remote work from Europe. That is one moment on a limited set of boards, but it suggests casting a wide net rather than waiting for one exact title.
A note for people switching in
If you are a backend engineer, family 1 is usually the shortest step. One or two shipped features that use an LLM API, with evaluations and sensible error handling, speak directly to what these postings ask for. If you work in DevOps, family 3 is the closer fit. Moving into family 2 usually takes more dedicated study of machine learning itself.
Whichever family you aim for, write your CV in the language of the posting. If the description says "evals" and "retrieval", describe your project in those words rather than a vague "worked with AI".
In Hot Jobs you can pick AI Engineer and Machine Learning as roles, alongside Backend or Data Scientist, and get new postings from company career pages in Telegram. €3 once, no subscription. Connect the bot