confusing questions in tech straight away is:
Each are six-figure jobs, but when you select the flawed one, you can waste months of your profession learning the flawed skills and miss out on quality roles.
As a practising machine learning engineer, I would like to stipulate the important thing differences and similarities between the 2 roles, so you recognize exactly which path suits you best.
Let’s get into it!
What Is The Difference?
Being honest, the industry is moving so fast that these titles change definition every quarter.
Not to say that firms now put “AI” of their job description to make the role seem more prestigious, though you’ll more than likely be doing basic prompt engineering.
Nevertheless, let me explain the difference, as I even have seen firsthand and discussed with other respected practitioners in the sphere.
In a nutshell, an AI engineer is a software engineer who specialises within the use and integration of foundational GenAI models akin to Claude, GPT, BERT, and others. They don’t “construct” these models, but relatively use them to serve a certain purpose.
However, a machine learning engineer is someone who actually develops models from scratch or using basic libraries and builds full end-to-end systems around them.
These are mainly more traditional models like gradient boosted trees and neural networks, but they may also be GenAI models.
What I find funny about this naming convention, is that machine learning is definitely a subset of AI.
So an AI engineer is technically a GenAI engineer, if anything.
Alright, enough of me being pedantic, let’s explain them in additional detail.
AI Engineer
What’s it?
As I discussed, you have got to consider an AI engineer as a software engineer that has a speciality in AI, well, GenAI.
They mainly work with something called foundational models, that are huge neural networks trained on oceans of knowledge akin to text, images, videos, and audio.
These foundational models can do many tasks, like writing code, answering questions, and creating images. That’s why they’re foundational, as they’ll achieve this many things.
OpenAI’s ChatGPT is essentially the most famous foundational model you’re likely aware of.
AI engineers don’t train these models; they integrate them into traditional software products and workflows using APIs, self-hosting, etc.
For instance, they might embed a chatbot on a shopping website to assist customers find what they’re on the lookout for more quickly, or add a coding assistant in an IDE, like Cursor.
AI engineering is more product focussed, i.e. you wish to deploy something quickly after which refine later.
What do they use?
This role is evolving quite a bit, but typically, you would like good knowledge of all the most recent GenAI, LLM, and foundational model trends:
- Solid software engineering skills
- Python, SQL and backend languages like Java or GO are useful
- CI/CD
- Git and GitHub
- LLMs and transformers
- RAG
- Prompt engineering
- Foundational models
- High-quality tuning
- Model Context Protocol
Machine Learning Engineer
What’s it?
A machine learning engineer focuses on constructing machine learning models and deploying them into production systems. It originally got here from software engineering, but is now its own job.
The numerous distinction between machine learning engineers and AI engineers is that the previous builds algorithms from scratch that give attention to more specific tasks.
For instance, machine learning engineers would construct targeted advice systems, bank card fraud models and stock forecasting algorithms. These aren’t “foundational” and have a much narrower use case.
For machine learning engineering, you should know these algorithms at a sophisticated level, which requires strong maths skills in statistics, linear algebra, and calculus. This will not be necessarily true for an AI engineer.
Machine learning engineering is more model-focused: you create the model from scratch using available data, test it offline, and ship it if you end up comfortable with its performance.
There also exist further specialties throughout the machine learning engineer role, like:
- ML platform engineer
- ML hardware engineer
- ML solutions architect
Don’t worry about these when you are a beginner, as they’re pretty area of interest and only relevant after just a few years of experience in the sphere. I just desired to add these so you recognize the varied options on the market.
What do they use?
The tech stack for machine learning engineers is analogous to that of AI engineers, with greater emphasis on mathematical abilities.
- Python and SQL, nonetheless, some firms may require other languages. For instance, in my current role, Rust is required.
- Git and GitHub
- Bash and Zsh
- AWS, Azure or GCP
- Software engineering fundamentals akin to CI/CD, MLOps, and Docker.
- Excellent machine learning knowledge, ideally with a specialism in an area like forecasting, advice system or computer vision.
- Solid mathematical understanding of statistics, linear algebra and calculus.
Which One?
As you’ll be able to see the overlap between skills and work is fairly similar, particularly the foundational software engineering skills.
The foremost difference lies within the domain specific GenAI knowledge of AI engineers and the deeper mathematical and traditional machine learning knowledge of machine learning engineers.
So, the query stands.
Let’s break down some more logistical features to provide help to in your decision.
Background
The background for each jobs is analogous, typically requiring a master’s in a STEM subject and a few years of experience as either a software engineer or an information scientist.
AI engineering is barely easier to get into, as learning to work with foundational models is a quicker learning curve than understanding all of the mathematics behind machine learning.
Demand
Machine learning engineering is the more established role, but that’s mainly because foundational models haven’t existed for long, so the AI engineer role wasn’t required.
Nevertheless, as AI is now super popular, demand for AI engineers is skyrocketing. You do should be careful, though, because job titles on this industry are vague, and you should really read the job description to know the job you can be doing.
For instance, at my company, we technically have AI engineers, but they’re still named machine learning engineers. So, titles are type of erroneous.
Pay
In response to Levels.fyi, the median salary for a machine learning engineer is £105k (UK) and for an AI engineer is £75k (UK), but I believe this may grow in the long run.
Plus, as I just stated, many machine learning engineers are doing AI engineering work, so the salaries are hazy.
Final Alternative?
For my part, go along with what you think that you’ll prefer!
When you love maths and understanding how algorithms work under the hood, then machine learning engineering is your best bet.
When you don’t like research that much and wish to quickly ship products using the most recent AI tools, then AI engineering is for you!
Either way, each roles pay well and have excellent long-term profession prospects.
Nevertheless, suppose you’re feeling a stronger pull towards a profession as a machine learning engineer.
In that case, I like to recommend testing my last article, where I’m going step-by-step through how I’d develop into a successful machine learning engineer all once more.
See you there!
One other Thing!
Join my free newsletter where I share weekly suggestions, insights, and advice from my experience as a practising data scientist and machine learning engineer. Plus, as a subscriber, you’ll get my FREE Resume Template!
Dishing The Data
newsletter.egorhowell.com
