Resources/Job description templates/AI engineer
AI Engineer Job Description Template (Free, 2027)
A free AI engineer job description template for engineers who build products on large language models: prompts, retrieval, agents, evals and production code, plus how the role differs from ML and research engineers. Copy, download as Word or print.
- 6sections
- Freedownload
Free template
- Copy it, download it as a Word file or print it
- Duties, requirements and FAQ
- Written by recruiters
Quick answer: An AI engineer job description should say whether the person builds products on top of existing models (LLM apps, agents, retrieval) or trains and tunes models, which models and tools you use, how you measure quality, and the pay range. The title means different things at different companies, so the first paragraph should make clear which kind of AI engineer you want.
Use this template for applied AI engineer roles: engineers who build features and products on large language models and other AI models. If you need someone to train models from scratch, adjust it toward a machine learning engineer or research engineer role. For pay, BLS has no AI engineer figure, so see our AI engineer salary benchmarks from Levels.fyi data.
AI engineer job description
About the role
[Company] is hiring an AI engineer to build [product or feature] using [large language models / your models]. You will take ideas from prototype to production and own how well they work for [users].
Location: [on site, hybrid or remote]. Reports to: [title].
What you will do
- Build features on [OpenAI, Anthropic, Google, open weight models or your models] through APIs and SDKs
- Design prompts, retrieval (RAG), tool use and agent workflows
- Build evaluation sets and measure quality, cost and latency before and after changes
- Fine tune or adapt models when prompting is not enough
- Ship production code in [Python / TypeScript] with tests, logging and monitoring
- Handle data privacy, safety and failure cases
- Work with product, design and customers to decide what to build
What you bring
- [3 or more] years as a software engineer, with [1 or more] years building with machine learning or LLMs
- Strong [Python / TypeScript] and API design skills
- Hands on experience with LLM APIs, embeddings, vector databases and retrieval
- A habit of measuring results with evals rather than judging by eye
- Clear communication with non engineers
Nice to have
- Fine tuning, distillation or model serving experience
- Experience with agent frameworks and the Model Context Protocol (MCP)
- Background in [your domain, such as recruiting, finance or healthcare]
- Open source AI projects you can link to
Pay and benefits
Pay range: [$ to $] base [plus equity and bonus]. [Benefits.]
How to apply
Apply at [link]. Send a link to something you built with AI and tell us how you knew it was working.
Free from Work Insiders. Edit it to fit your team.
AI engineer, ML engineer or research engineer: which do you need?
A mismatched title is a common reason AI job ads pull in the wrong applicants. If the job is mostly building on vendor models, call it AI engineer and say so.
| Title | Main work | Typical background |
|---|---|---|
| AI engineer | Builds products on existing models: prompts, retrieval, agents, evals | Software engineering |
| Machine learning engineer | Trains, tunes and serves models; builds data and training pipelines | Software plus ML |
| Research engineer | Implements and scales new model research | ML research, often a graduate degree |
| Forward deployed engineer | Deploys AI products inside customer environments | Software plus customer facing work |
What should an AI engineer job ad avoid?
Long lists of every framework, unrealistic years of experience for tools only a few years old, and vague claims about working on AI. Name the product, the models and the quality bar. Our directory compares technical assessments that test applied AI skills.
Frequently asked questions
What does an AI engineer do?
Builds software products and features on top of AI models, mainly large language models: prompts, retrieval, agents, evaluation and the production code around them.
What is the difference between an AI engineer and a machine learning engineer?
AI engineers usually build on existing models; machine learning engineers usually train, tune and serve models. Many companies use the titles loosely, so describe the work.
How much should I pay an AI engineer?
BLS has no AI engineer category. Levels.fyi data on our salary page show a $160,000 median total compensation for the AI engineer job title and $282,950 for machine learning engineers, as of October 2026. See AI engineer salary.
Does an AI engineer need a PhD?
No. Applied AI engineer roles usually hire strong software engineers with hands on LLM experience. Research roles are where graduate degrees matter.
Sources
Written by recruiters at Work Insiders. Updated November 4, 2026. How we make money.