AI Engineer
What the work looks like, the stacks behind the title, what each level means and the training that helps, then how interviews run and how to prepare for each step.
What an AI engineer actually does
An engineer who builds product features on top of LLMs and foundation models, and makes them reliable, safe and affordable in production.
An AI engineer builds applications and product features on top of large language models and other foundation models. They connect the models to the company's data and systems, design the prompts, build retrieval with RAG and write the code around the model.
Most of the work is about quality in production. AI engineers build evaluation sets, measure answers, add guardrails, trace what the model does and control the cost and the latency of each call.
They choose between model APIs and open models, and sometimes fine-tune a model for a specific task. More and more teams also work with AI coding tools such as Claude Code, Cursor or GitHub Copilot, and expect engineers to use them with judgment.
- Building or improving an LLM feature in the product
- Running evaluation sets after a prompt or model change
- Fixing wrong answers reported by users
- A review of model costs and latency
- Code reviews, tests and documentation
The title spread from 2023, as companies started to build products on top of LLMs, and the AI Engineer Summit gave the community a name and a place to meet.
Some companies still use the title AI Engineer for machine learning engineers who train their own models. The job ad, and the share of work on model APIs versus model training, show which role it really is.
The same title, very different stacks
Most AI engineering searches are defined first by the part of the LLM application the team focuses on, then by the ground every AI engineer shares.
What each level means in practice
Titles vary from one company to the next. What changes from one level to the next is the scope you own and the decisions you take.
- Works on LLM features designed by a senior engineer
- Learns the models, the data and the evaluation tools
- Writes tested code with code review
- Builds and ships LLM features on their own
- Runs evaluations and fixes quality issues
- Works directly with product and data teams
- Designs LLM systems for a product area
- Takes choices on models, retrieval and cost
- Mentors other engineers and leads code reviews
- Sets the AI architecture across teams
- Leads choices on models, vendors and standards
- Writes the design documents others build on
The paths into the role, and what really counts
Companies look first at LLM features you have shipped and how you measured their quality. Education and certifications help most early in a career, or to change field.
- Engineering school or master's in computer scienceThe most common path in France, and often expected by large companies.
- University degree in computer scienceThe most common path in the US, at bachelor's or master's level.
- Backend or full-stack engineering firstMany AI engineers start as backend or full-stack engineers before moving to LLM applications.
- ML engineering or data science firstML engineers and data scientists also move into the role, with a strong base in models.
Our view: for this role, an LLM feature you shipped and can walk through, from the prompt to the evaluation results, often says more than any certificate.
A conversation with a senior tech recruiter
Your first interview is with one of our senior tech recruiters, who knows the role, the stacks and the levels in this guide. The conversation covers your own work through real examples, and the technical exercises come later, with the company. Before the call, we send you the six themes we will cover so you can prepare your examples.
- What you owned yourselfThe LLM features or AI applications you were responsible for, and your own part in the team's work.
- Your environment and its constraintsData access, quality needs, latency, budget or security rules: what shaped your technical choices.
- The decisions you madeWhich model, which retrieval and which guardrails, the options you compared and the reasons for your choice.
- The impact of your workWhat changed thanks to your work, with numbers when you have them: answer quality, adoption, cost or latency.
- How you work with other teamsYour work with product, data and security teams, and how you settle a disagreement.
- What you want nextThe kind of AI product, users and scope you are looking for in your next role.
- LLM features in production
- How quality is evaluated
- Cost and latency trade‑offs
The same ground for every candidate. For a given role, every candidate answers the same final questions, built with the hiring manager around what the role really needs.
Ready for the next steps. After our call, we tell you how the company runs its interviews, who you will meet at each step and what each step is meant to assess.
What the company will assess, step by step
Each company runs its own process, and we walk you through its steps before you start. AI engineering processes often combine the steps below, and the depth expected at each step grows with your level.
- A coding exerciseLive coding or a take-home, often a small LLM feature built with a model API, with tests or evaluations.
- An LLM system designA system to design, such as a RAG assistant on company documents, with retrieval, evaluation, guardrails and cost.
- A deep dive on your past workOne LLM feature you shipped, discussed in depth: the context, your choices and what you would change today.
- Meeting the manager and the teamHow you work, how you take decisions with others and what the team expects in your first months.
- ConfirmedYou solve the exercise cleanly, test the answers of the model and explain each choice in simple words.
- SeniorYou lead the design, weigh quality, cost and latency, and point out risks and limits before you are asked.
- Staff and PrincipalYou run most of the interview as a peer, think about the whole AI platform and connect the design to the needs of several teams.
How to prepare as an AI engineer
Practical advice for each step of an AI engineering process, from the way you introduce yourself to the questions you ask at the end of each interview.
Start with the LLM features you shipped, their users and your part in them. Mention your stack in one sentence, then close with what you want next.
For each one, prepare the use case, the model, your decision, the options you set aside, the result and what you would change today. Say "we" for the team and "I" for your own part.
Interviewers often ask how you knew the answers were good. Prepare your evaluation sets, your metrics and one change you made because of them.
In an LLM system design, start with the users and the data, then the retrieval, the model, the guardrails and the evaluation, in that order.
Prepare the cost per request, the latency and the volume of the features you shipped, and how you reduced them.
Rules on AI coding assistants differ from one company to the next, and sometimes from one step to the next. Ask before each step, and when they are allowed, explain what you ask the tool and how you check its output.
- Walk us through an LLM feature you shipped, from the prompt to production.
- How do you evaluate the quality of a RAG system?
- How do you reduce the cost and the latency of an LLM feature?
- When would you fine-tune a model instead of using prompts and retrieval?
- Which models and providers does the team use today?
- How does the team evaluate quality before and after a release?
- Who reviews the security and the risks of AI features?
- How does the team follow model costs in production?
- How do AI engineers and ML engineers share the work?
- How does the team use AI coding tools day to day?
Where AI engineers come from, and where they go
The roles people most often hold before this one, and the ones they most often move to.