AI Engineer Jobs: What You Do, Which Degree to Get and How to Get Hired

AI engineer jobs are software jobs built around machine learning. You write code that trains, connects and monitors models so a product can use them. BLS counts AI engineers with software developers, so these figures cover that wider group: a May 2025 median of $135,980. A bachelor’s is the usual way in, and a master’s is common for research-heavy roles.

What Is an AI Engineer?

An AI engineer is a software engineer whose product depends on a machine learning model. A recommendation feature, a document reader, a support chatbot or a fraud check all need someone to get the model running inside real software and keep it running.

That is different from the job most people picture when they hear “artificial intelligence.” A degree in AI is a course of study. An AI engineer is a person with a job title. You can hold the job with several kinds of degrees, and you can hold an AI degree and work in another job entirely.

The title also has no official federal category. BLS counts these workers as software developers, and we say so each time we use a number on this page.

What Do AI Engineers Do Day to Day?

BLS describes software developers as people who analyze users’ needs, design and build software to meet them, plan how the pieces of a system fit together, and keep programs working through maintenance and testing. An AI engineer does that with a model in the middle.

A normal week often mixes these tasks:

• Cleaning and shaping data so a model can learn from it.
• Training a model, or adjusting one a vendor or research team built.
• Wrapping the model in an interface other software can call.
• Testing results, then checking them again after the data changes.
• Watching cost and speed, because a model that works but runs slowly or costs too much will not ship.
• Explaining limits to managers who want the feature to do more than it can.

Notice how much of that is ordinary software work. If you like writing and fixing code, you will spend most of your time doing it. If you want to spend your days inventing new model designs, that is closer to research, and we cover it below.

How Does AI Engineering Differ From AI Research?

Employers use “AI engineer” for two quite different jobs, and the posting usually tells you which one.

Building and shipping. You use existing models and tools to make a product work. BLS lists a bachelor’s as the typical entry education for this group.

Research and design. You develop new methods and test them. BLS files this under computer and information research scientists, where it lists a master’s as typical. That occupation is much smaller, at 37,200 workers in May 2025, with a median of $140,300. BLS projects 22 percent growth for it from 2025 to 2035.

The medians are only a few thousand dollars apart, so pay is not what separates the two. What differs is the education that employers expect and how many openings exist. Read each posting’s required skills to see which side it sits on.

Which AI Degree Level Do Employers Expect?

There is no single answer, because the job covers both kinds of work. Here is how the levels line up with what the data shows.

Bachelor’s. BLS lists a bachelor’s in computer and information technology or a related field as typical for software developers. An AI or computer science bachelor’s fits building roles. Our online bachelor’s in AI ranking sorts these by price, format and graduation rate. The bachelor’s totals on our list run from $44,856 to $67,800 for 120 credits.

Master’s. This is where most online AI degrees sit. It fits two kinds of people: those who want research-style roles, and those who already have a bachelor’s and want AI skills. The median total tuition for an online AI master’s on our list is $32,463, and the range is wide. Browse the online master’s in artificial intelligence ranking to compare.

Graduate certificate. Short programs fit working engineers who need the tools, not a new degree. Carnegie Mellon’s online AI Engineering graduate certificate totals $16,968 and lists 12 months. Penn State’s totals $12,576.

Doctorate. For research jobs at the frontier. Few engineering roles ask for one.

If you want programs built around engineering, not theory, the online AI engineering degrees ranking covers them at every level.

Which Kinds of AI Programs Fit This Job?

Titles vary, and the title alone tells you little. Three kinds of program prepare you well for AI engineer jobs.

AI engineering programs. Western Governors University’s online BS in AI Engineering ends in a capstone that deploys an AI system with industry-standard MLOps tools, and it lists about 24 months. Penn State’s online master’s has an AI Engineering option that ends in an applied project paper.

Applied AI programs. Boston University’s online MS in Software Engineering for AI totals $25,000 and has no GRE requirement. Saint Louis University’s applied AI master’s totals $24,300 and does not require a STEM degree, which helps if you are changing careers.

General AI master’s programs. The University of Texas at Austin’s online MS in Artificial Intelligence totals $9,990 for 30 credits, and its courses include deep learning, natural language processing and reinforcement learning.

Our cheapest online AI degrees list sorts every level by price, which is the quickest way to see how far a budget stretches.

What Skills and Tools Do You Need?

Admission pages are the best public guide, because schools only ask for what their first courses assume. Johns Hopkins lists calculus, linear algebra, probability and statistics, and Python. Carnegie Mellon’s certificate asks for linear algebra, calculus, statistics and probability, plus Python or other programming experience.

Put that together and you get a short skills list:

• Python, the language that most admission pages name. Some also accept R.
• Data structures and algorithms.
• Linear algebra and calculus, mostly so you can follow how models learn.
• Probability and statistics, so you can tell a real result from luck.
• SQL and basic data handling.
• Git, the command line and tests, which no admission page lists but every team expects.
• Cloud basics. Some programs mention cloud usage costs for assignments, so budget for it.

Our AI degree requirements guide goes through what schools ask for in more detail.

How Do You Get Hired as an AI Engineer?

Here is a path that works for most people, in order.

1. Check your foundations. Be honest about Python, statistics and linear algebra. If one is weak, fix it first with a short course.
2. Choose a degree level that fits your goal. Building products points to a bachelor’s or applied master’s. Research points to a master’s or higher.
3. Pick a program with applied work. Capstones give you something real to discuss. Ask what the capstone produces and whether you keep the code.
4. Build two or three projects outside class. Deploy a model that someone else can use. Write down what broke and how you fixed it.
5. Search under several titles. Use machine learning engineer, applied scientist, data engineer and software engineer with ML in the posting.
6. Apply for software roles too. Many teams hire general software engineers, then move them onto AI work.
7. Prepare for two interview types. Expect coding questions and a conversation about a project you shipped.

Our guide to becoming an AI engineer covers prerequisites, bridge programs and timelines step by step.

What Do AI Engineer Jobs Pay?

BLS has no separate AI engineer line. We use software developers, the group BLS counts these workers in, so these figures cover that wider group and not AI work alone.

The May 2025 national median is $135,980 a year, or $65.38 an hour, across 1,687,890 workers. The mean is $148,100. The middle half earns $105,210 to $171,980, the lowest tenth earns $82,460 or less and the highest tenth earns $214,670 or more.

Location moves the number most. These states have the highest annual medians:

• California: $174,410, with 284,390 workers.
• Washington: $166,540, with 107,030 workers.
• New York: $166,180, with 113,510 workers.
• Massachusetts: $165,210, with 48,190 workers.
• Oregon: $142,720, with 21,830 workers.

At the other end, Mississippi’s median is $95,330 and South Dakota’s is $96,310. Texas has the second-most jobs, 163,880, with a median of $132,150. Our AI engineer salary page has the percentiles, more states and the cost of a degree set against pay.

Are AI Engineer Jobs in Demand?

BLS projects employment of software developers to grow 10 percent from 2025 to 2035, which it calls much faster than average. It expects about 185,400 new jobs on a base of 1,905,400, and about 106,100 openings a year once people who leave the field are counted.

Those figures cover all software developers, not AI work alone. We can’t say how many of the openings involve machine learning.

Projections are estimates, and BLS revises them each cycle. Entry-level competition can run higher than the growth rate suggests, because many people are applying for the same junior roles. A portfolio and a clear specialty help you stand out from the crowd of applicants.

Can You Work Remotely as an AI Engineer?

Sometimes. The data we use does not measure remote work, so we can’t tell you how common it is. What we can say is that software work does not require a particular building, and you will find remote and hybrid postings on job boards.

Three things to check before you count on it. Some employers set pay by where you live, so the table above is a guide, not a promise. Some require occasional office visits. And junior roles are more likely to be on site, because teams want to coach newer hires.

Can You Get Hired With an Online AI Degree?

Yes. A degree earned online is still a degree from the school that issued it. What matters is the school’s accreditation, what the program taught, and what you can show.

Check four things before you enroll:

• The school is regionally accredited.
• The curriculum covers machine learning and how to deploy a model, not only theory.
• The program ends in a capstone or project you can show.
• Any in-person requirement fits your life. A few programs, such as Duke’s, include onsite time.

Online programs vary in format. Some run live sessions at set times, and others let you work at your own pace. If you are working full time, that choice can decide whether you finish.

What Mistakes Should You Avoid?

Three show up often. The first is choosing a program by its title alone. Two degrees both named “AI” can teach very different things, so read the course list and look for deployment, not just theory.

The second is skipping prerequisites. Carnegie Mellon’s certificate, for example, expects linear algebra, calculus, statistics and probability before you start. Applying without them can cost a fee and a term.

The third is treating the degree as the whole plan. Employers still ask what you built. Start your projects while you study, and keep the code somewhere a hiring manager can read it.

What Should You Do Next?

Pick the one step that matches where you are today. If you already code, look at a graduate certificate or an applied master’s. If you do not, start with a bachelor’s or a bridge course. If cost is your main limit, open the cheapest list and sort by level.

Then talk to two people who hold the job. Ask what they were hired as, what they studied and what they would skip. Their answers will tell you more about your local market than a national survey can.