What is AI engineering?
AI engineering is the discipline, and the job, of building applications on top of readily available foundation models: the prompts, context, evaluation and fine-tuning around the model. Shawn “swyx” Wang popularised the role in June 2023, and Chip Huyen’s 2024 book defines the field. Its subject is AI inside the product. Using AI to write code is a different subject.
Origin Shawn “swyx” Wang popularised the role in “The Rise of the AI Engineer” (opens in a new tab) on 30 June 2023, noting that he was “calling attention to this trend rather than starting it”, and Chip Huyen’s book (opens in a new tab), first released on 4 December 2024, defines the field as “the process of building applications with readily available foundation models”.
- In Polish
- Inżynieria AI
- Closest ladder level
- All levels, free guide
- Also searched as
- AI engineer, LLM engineer, foundation model engineer
Then $19.99 a month or $99.99 a year. Cancel from your account page. 30-day refund on a first purchase.
Why it matters.
Two readers use this phrase for different jobs. An AI engineer ships features whose product is model behaviour and checks them with evaluation pipelines. Someone who uses coding agents to write ordinary software is doing agentic engineering, whatever the job title says. A job post, policy or budget line that says “AI engineering” without saying which sense it means will be answered by the wrong reader. Some pages on this site use the phrase in the second sense; this page uses the first.
AI engineering vs vibe coding
They sit on different axes: vibe coding is about how code gets written, and AI engineering is about what gets built, so neither is the opposite of the other.
| Compared on | Vibe coding | AI engineering |
|---|---|---|
| What it names | How code gets written: accepting AI-written code without reading it | What gets built: applications on foundation models. Also a job role |
| Who writes the code | The model | Human engineers; the term says nothing about AI writing the code |
| Who checks it | Nobody reads the diff; the human checks behaviour by eye | The engineers, through evaluation pipelines that test the product |
| What stops the loop | “it mostly works” (Karpathy) | Nothing: it is a profession, not a loop |
| Where the name came from | Andrej Karpathy, post on X, 2 Feb 2025 | swyx popularised the role, 30 Jun 2023; Chip Huyen’s book, 4 Dec 2024 |
| Where it breaks | Past “throwaway weekend projects”, the scope Karpathy gave it himself | When the phrase is read in the other sense: it also gets used for AI-assisted coding |
AI engineering vs agentic engineering
Both names put “AI” and “engineering” side by side, and they describe different jobs: one builds AI into a product, the other uses AI to build any product.
| Compared on | AI engineering | Agentic engineering |
|---|---|---|
| Where the AI sits | Inside the product the team ships | In the toolchain: agents that write and test the software |
| What gets built | AI-powered applications and features | Any software |
| Who checks the result | Engineers, through evaluation pipelines | Tests and evals, with the human owning quality |
| Who popularised the name | swyx, 30 Jun 2023 | Andrej Karpathy, 4 Feb 2026 |
Questions about AI engineering.
How is AI engineering different from vibe coding?
They answer different questions. AI engineering is about what gets built: software whose product is model behaviour. Vibe coding is about how code gets written: an LLM writes it and nobody reviews it, a term Andrej Karpathy introduced on 2 February 2025. An AI engineer may or may not use coding agents, and a vibe coder may build an app with no AI inside it. No source compares the two; this is our reading.
Is AI engineering the same as agentic engineering?
No. AI engineering puts AI inside the product. Agentic engineering uses coding agents to build any software without lowering the quality bar, a name Andrej Karpathy popularised on 4 February 2026. The two overlap only when an AI engineer uses coding agents. No source contrasts them directly, so the split is our reading, and some pages on this site use “AI engineering” in the agentic sense.
What is the difference between an AI engineer and an ML engineer?
AI engineering grew out of ML engineering. Chip Huyen writes: “While AI engineering is a new term, it evolved out of ML engineering”. The starting point differs: AI engineers build on foundation models that already exist, and swyx put the hiring logic in 2023 as “When it comes to shipping AI products, you want engineers, not researchers.”
Does “AI engineering” ever mean using AI to write code?
Loosely, yes. Every’s Kieran Klaassen wrote on 18 August 2025: “Typical AI engineering is about short-term gains. You prompt, it codes, you ship.” In that sense the phrase means AI-assisted coding. Our reading is that the usual meaning is still building applications on foundation models, so a policy or job post should say which sense it uses.
The vocabulary of AI-driven development
Every name below is defined against vibe coding: who writes the code, who checks it, and what stops the loop.
Read the long-form guide: agentic engineering vs vibe coding
Sources.
The primary sources outside this site that this page relies on.
- The Rise of the AI Engineer (opens in a new tab) Shawn “swyx” Wang, Latent Space
- AI Engineering: Building Applications with Foundation Models (opens in a new tab) Chip Huyen, O’Reilly
- AI Engineering: chapter summaries (opens in a new tab) Chip Huyen, GitHub
- My AI Had Already Fixed the Code Before I Saw It (opens in a new tab) Kieran Klaassen, Every
Keep reading.
The guides that go deeper, and the terms and comparisons next to this one.
In the docs
Related terms
Comparisons
Read the guides in the same words.
Open every guide with the 7-day free trial. Each term is defined once here, and the guides use it the same way.
Then $19.99 a month or $99.99 a year. Cancel from your account page. 30-day refund on a first purchase.