The machine learning job market is hot, and only getting hotter. Focused on training computers to learn over time and produce more and more accurate models that predict future occurrences, machine learning is increasingly being deployed in industries like finance, healthcare, and manufacturing. Companies in these industries are eager to hire those with machine learning skills who can help them build better products, streamline logistics, and increase their bottom line.
The most popular machine learning role that companies recruit for is the machine learning engineer. In 2019, Indeed reported that postings for machine learning engineer jobs accounted for 179 of every one million job postings in 2018, with the total number of machine learning engineer postings growing by nearly 350% between 2015 and 2018.
Perhaps you’ve already heard about this booming job market, or maybe you’ve heard about the impressive salaries offered even to entry-level machine learning engineers. Maybe you’re attracted by the opportunity to play a role in cutting-edge technologies that are increasingly impacting people’s lives, or maybe you’re just figuring out your next move and mulling over your options. Whatever the case, if you’re here you’re likely interested in learning how to become a machine learning engineer. We’re here for you: in this article, we’ll give you the low-down on machine learning engineering and lay out the steps you can take to become a machine learning engineer.
First thing first: what exactly is a machine learning engineer?
What is a machine learning engineer?
A machine learning engineer is responsible for designing, developing, and shipping machine learning models — and then maintaining them once they are shipped and deployed to a live product. An ML engineer might work on a team focused on building a particular product or feature, or they might work as a generalist and move between a variety of different types of projects. In addition, machine learning engineers can be responsible for managing data systems and undertaking various kinds of data analysis, such as analyzing data sets from A/B tests on a machine learning algorithm. Sometimes, a machine learning engineer might also be tasked with data visualization tasks.
All this doesn’t take place on its own. ML engineers use their machine learning skills to solve real-world problems, so their efforts can have a measurable effect on the lives of real people. As an ML engineer, you can work in breaking-edge areas like natural language processing (NLP), which focuses on teaching computers to understand and use human language, or deep learning, a kind of machine learning in which a neural network — is used to optimize performance. Other exciting areas include computer vision, which focuses on teaching computers to be able to perceive, analyze, and react to visual stimuli — with important applications including autonomous vehicles or medical imaging diagnostics.
A machine learning engineer doesn’t work alone: often you can expect to work in close collaboration on cross-functional teams. On a typical day, you might interface with a data scientist, a software developer, or even a product manager.
How much does a machine learning engineer make?
One of the biggest draws for the profession is the machine learning engineer salary. After all, ML engineering is not only one of the fastest growing job areas, but one of the best compensated. LinkedIn puts the 2022 salary range at $72,600-$170,000, while Indeed puts the average machine learning engineer salary at $123,408.
This means that an average machine learning engineer earns over double the $58,260 that the U.S. Bureau of Labor Statistics estimates the average American earned in 2021. And machine learning engineer salaries remain attractive even when considered against other technical professions. A data analyst, for example, makes on average just $69,973 per year as a base salary, and an entry-level cyber security analyst takes home $73,242 yearly — both well below the six-figure salary most machine learning engineers can expect.
What kinds of skills does a machine learning engineer need?
Companies investing money for nothing, of course. To earn such a high salary, a machine learning engineer must have a substantial arsenal of skills. While the exact skills and competencies will vary depending on the role and the nature of the projects, for an applicant for a machine learning job to have any chance of landing an interview, they need to be able to demonstrate a sufficient background in key areas of mathematics, including probability, statistics, and linear algebra.
Engineers also need to have significant coding ability to write machine learning algorithms. Many companies will expect applicants to be fluent in at least two coding languages such as C++, Python, Java, or R programming language. This will be assessed either from the projects that populate an applicant’s portfolio (more on this later!) or through a coding challenge.
In addition to just knowing their way around a programming language (or several) ML engineers also need to have experience writing machine learning algorithms and building machine learning models on platforms like PyTorch and TensorFlow. Most machine learning engineers also have experience with cloud computing platforms like Azure and Amazon Web Services (AWS) — which allow for easier and more efficient analysis of big data sets — and data pipeline tools like Apache Bearn.
While a machine learning engineer will leave school with a broad understanding of the different subdisciplines of machine learning, often — and especially in more senior roles — an ML engineer will specialize in natural language processing, deep learning, or computer vision.
In addition to these skills, machine learning engineers establish crucial subject knowledge and industry experience. A machine learning model works very differently in the lab than it does in the real world, so machine learning engineers need to know their way around the industries they’re working in. Take, for example, Derek Driggs, a researcher at Cambridge who realized that his machine learning model for diagnosing COVID tended to give false positives for seriously ill patients because it had learned to diagnose solely based on whether a patient was seated (and so likely healthier) or lying down (and so likely iller, regardless of their illness).
Of course, all the technical artificial intelligence and machine learning skills won’t mean a thing if they don’t rest on a foundation of crucial soft skills. Machine learning engineers need to be able to collaborate on cross-functional teams, and that means having experience communicating with people of different backgrounds and expertise. As we noted above, on a given day an ML engineer might meet with a data scientist, a data engineer, and a product manager and be expected to discuss their progress on a new ML model, explain the issues with an existing ML model, or communicate how their work fits in with other business initiatives — all in a way that will be understandable to each stakeholder. Machine learning engineers must also have project management skills to keep their work on track through the version control software Github and creative thinking skills to be able to ideate novel solutions to problems.
What kind of education does a machine learning engineer need?
In general, employers looking for junior machine learning engineers will seek out applicants who have a bachelor’s degree in machine learning, computer science, or artificial intelligence and who can also demonstrate some on-the-job experience, such as a relevant internship. Some companies will interview applicants who have bachelor’s degrees in other fields but have completed a data science or machine learning bootcamp or online short course in ML, though here it becomes especially crucial to have a strong portfolio of work.
For more senior machine learning engineer roles, companies will usually require that applicants have a master’s degree or even a Ph.D. in machine learning, computer science, or artificial intelligence, plus significant relevant work experience.
But what exactly does each educational path entail? We’ll dive into the specifics now.
Machine Learning Bachelor’s Degrees
Students looking to study machine learning as part of a Bachelor of Arts (BA) or Bachelor of Science (BS) course of study have options when it comes to picking a major. While most instruction in machine learning is still taking place through artificial intelligence or machine learning specializations added to traditional computer science majors, standalone AI and ML majors are becoming increasingly popular.
Regardless of home department, the curriculum will be largely the same. Students will take foundational courses in math, statistics (e.g. Multivariable Calculus, Linear Algebra, Probability), and computer science (e.g. Computing Systems I&II, Principles of Functional Programming Database Systems, Principles of Software Engineering), as well as machine learning-specific introductory courses and electives (e.g. Deep Learning, Natural Language Processing, Computer Vision).
Many programs also require a capstone project to be completed during your final year of study. Before that, students will generally take a course in AI ethics and policy, a writing and communication course, as well as distributional requirements in the social sciences and humanities. Often, undergrads will also have the opportunity to assist their professors in cutting-edge machine learning research.
Machine Learning Bootcamps
A machine learning bootcamp offers a short and flexible course of study to equip students with the most important machine learning skills. Due to their highly technical nature, shortened instruction schedule, and frequent reliance on self-study, many machine learning bootcamps set formal or informal prerequisites for applicants to ensure that those they admit will succeed.
While each program is different, most are looking for applicants who already have significant experience in either computer programming or mathematics — and preferably both. While formal work experience and a bachelor’s degree in computer science or a data-heavy STEM field will be the most obvious indications of these skill sets, many machine learning bootcamps are also open to applicants who can demonstrate that they’ve developed these skills through self-study and independent projects.
All bootcamps are built differently, but a student taking a machine learning bootcamp can generally expect to focus on:
1) learning the basic theory behind artificial intelligence and machine learning,
2) building key skills and familiarity with the software that a machine learning engineer works with every day, and
3) completing a capstone project that will show prospective employers that the student can apply everything they’ve learned.
Machine Learning Master’s Degrees
A master’s in machine learning is an advanced degree, usually offered by a computer science department, that’s intended to familiarize students with the landscape of machine learning and give students the skills and other practical knowledge they need to undertake doctoral research or build and deploy artificial intelligence solutions in the industry.
While the majority of entering students will have completed bachelor’s degrees in computer science, master’s programs also accept students with backgrounds in other STEM fields, provided they have the requisite mathematics and programming skills, or even from the social sciences and humanities.
Certain lucrative positions (e.g. senior ML engineer or deep learning engineer roles) in Big Tech and other leading companies are usually so competitive that only those with master’s degrees will get interviews, so as machine learning engineers move through their careers, they will often pause for master’s study before resuming working at a higher compensation level.
Machine Learning PhDs
Those looking for an opportunity to engage in more extensive research in machine learning will often enter Ph.D. programs within computer science departments. These kinds of programs are increasingly gaining popularity. Indeed, the percentage of graduating CS PhDs who specialize in artificial intelligence and machine learning has grown dramatically over the last ten years: Stanford’s AI Index reports that such students made up over 20% of all CS Ph.D. students in 2019, up 8.6% from 2010 and dwarfing all other CS specializations.
At the same time, pursuing a Ph.D. in artificial intelligence remains an exclusive club. With the rigorous requirements and lengthy time-to-degree (generally 4-5 years), it’s no surprise that the same survey counted less than 300 new Ph.D. students specializing in AI or ML graduating from American computer science departments in 2019.
In the US, students showing promise are frequently admitted to Ph.D. programs without a master’s degree, with the understanding that they will complete a master’s course of study (and the requisite comprehensive exams) before being allowed to begin doctoral research (a process usually known as “qualifying”).
While the primary focus in Ph.D. programs is on research, more and more newly-minted machine learning PhDs are entering the industry, working as high-level machine learning engineers in artificial intelligence and machine learning labs like Google’s DeepMind and Meta AI.
How can you start on the path to becoming a machine learning engineer?
So now you have some visibility into what a machine learning engineer does, the salaries available, and crucial skills and the educational paths that will help you build them. What are the next steps?
Step One: Research programs.
First, you’ll want to narrow down the plethora of programs to find a couple that would be a good fit. Beyond the obvious choices — like whether you already have a bachelor’s degree, and so should look at a bootcamp or master’s program, or should start instead with a four-year bachelor’s program — you’ll want to make choices about:
Flexibility: Do you want to study part-time or full-time?
Modality: Do you want to study online, on-campus, or in a hybrid capacity?
Location: Do you want to study close to home, or would you like to explore a new city?
Reputation: Does a school’s prestige matter to you — and, importantly — will it matter to your prospective future employers?
Cost: What sources will you use to pay for school, and what is your budget?
As you research, we’re here to help. With so many factors at play, it’s easy to get overwhelmed when deciding on a program. Ultimately, the decision will be yours: you’re the expert on your current situation and your goals for the future. But to make your research and decision easier, we’ve assembled some expert-vetted recommendations to help you find what’s right for you.
Step Two: Apply to those that fit your background and goals.
After you’ve settled on some schools that you think would be a good fit, you’ll need to apply for them. To do this, you’ll need to get to know each school’s online portal and figure out what they require applicants to provide. Do they want you to write a personal statement? Provide GPAs, transcripts, or test scores? Are there set deadlines, or do the schools have rolling admissions? Again, we’re here to help. Head to our Education portal for the latest application details, as well as tricks and tips to make your application stand out from the crowd.
Step Three: Make the most of your time in school.
You’ve applied to schools, gotten admitted somewhere, and enrolled. Now you need to make the most of your time there. Already at the beginning of your machine learning course of study, you should be thinking about your exit: how can what you do today set you up for tomorrow? Here are some ideas:
Adopt a “yes” mentality: Accept any opportunities that come your way.
Network like it’s your job: School is a great time to build relationships that can help you progress in both your personal and professional lives.
Study hard: When there are thousands of applications for a single machine learning engineer job, those with low GPAs will have a much harder time getting an interview. Keep your nose in the books to keep yourself in the running later on.
Practice, practice, practice: While you’re at school, you’ll want to build a portfolio of projects that you can show prospective employers. Get as many internships as you can, work hard on your capstone, and save some time for personal passion projects along the way.
Step Four: Put yourself out there.
If you’ve made the most of your time in school, you’ll be in a good position when it comes time to start applying to programs. Your time spent networking will help land you referrals that will make sure your resume gets read. Your time spent working on machine learning projects will leave you with a stellar portfolio to show off in your interview. And an earnest and proactive attitude will ensure you have plenty to feature on your resume. If you need any tips on how to put it together, check out our machine learning resume guide.
Step Five: Be persistent.
Of course, you can only do so much. Often, there will be several great candidates for a machine learning job, with the deciding factor simply being which one is the best fit on the team and at the company. At a certain point, you can’t control this, so apply widely and be patient: if you’ve followed these steps, something great will come your way.
Step Six: Profit.
It will be a glorious day when you get the call that informs you that you’ve landed your first machine learning job. Bask in the satisfaction of a job well done, do a great job once you start, and think about your next move and how you can get there. If you need any advice along the way, we’ll be here, ready to help you move forward. To keep in touch and stay informed about the latest developments in the field, sign up for our newsletter.