Getty Images/iStockphoto
How to get started with machine learning
Machine learning roles are rapidly evolving and require a diverse range of skills. Looking to join the field? Start by exploring job responsibilities and required experience.
The meteoric rise of machine learning and AI is creating new employment opportunities for enterprise technology professionals. Some of these jobs are extensions of existing or traditional technology roles, while others are newly envisioned to meet the unique needs of building ML and AI.
According to the U.S. Bureau of Labor Statistics, careers in computer and information research -- a category that includes areas like software, data and machine learning (ML) -- are expected to grow by 23% between 2022 and 2032, a growth rate much faster than average. To capitalize on this upward trend, job seekers need to understand what types of ML opportunities are available, what skills they require, and the best ways to advance their careers.
Machine learning job skills
Requirements for ML jobs vary, but all are founded on two core skills: computer programming and mathematics.
Whether building or training a new ML model or using pretrained models for an analytics project, ML professionals typically rely on coding skills in one or more major programming languages. These often include Python and R, as well as Java and numerous others depending on the project and goals. ML professionals also use a range of software frameworks and libraries specialized for AI projects and data analysis, such as Pandas, NumPy and TensorFlow.
ML roles also often require a practical working knowledge of algorithms and mathematical skills. Key areas to focus on include probability, statistics and linear algebra, as well as calculus for the most complex data relationships, calculations and transformations.
Beyond this foundation of programming and mathematics, specific skills vary for each ML role based on business needs, goals and resources. Compared with more traditional areas of software engineering, ML is still an evolving field; roles are likely to change, merge or even diverge over time as practices become better established and defined.
Types of machine learning jobs
Although many of the technologies and concepts underpinning ML are decades old, the practical implementations of the technologies needed to build and operate ML and AI at scale -- such as high-performance computing, global networking and vast data availability -- are more recent. Businesses racing to embrace ML and AI projects need capable technology professionals to ensure those projects' success.
Machine learning engineers
ML engineers typically focus on the design, development, deployment and maintenance of ML models and their underlying algorithms. ML engineers are mainly software developers responsible for model creation using existing and newly created ML libraries. Their responsibilities also often extend to more traditional IT tasks, such as resource provisioning, as well as model deployment, monitoring and support. They frequently work with other ML and AI roles.
Beyond programming skills and a clear knowledge of probability and statistics, an ML engineer typically uses skills including data modeling, software systems design and software modeling for ML algorithms. ML engineers also need a practical knowledge of IT hardware and infrastructure.
Data scientists
Data science specialists gather, manage and use the data needed to train and operate ML and AI systems. They understand where data comes from and make sure it is current, complete and correct. They also ensure that data is securely stored and retained in accordance with retention policies and analyze how it can be used to train ML models. Data scientists often work with business specialists for analytical tasks and ML engineers who use data for model training.
Data scientists should have some programming knowledge, but their major emphasis is on mathematics, including probability, statistics, algebra and calculus. Data scientists also need skills related to data management and manipulation, including database management, data modeling and data visualization. Many data scientists also possess some expertise in ML algorithms, models and related specializations like deep learning and large language models (LLMs).
Machine learning designers and trainers
ML designers and trainers -- sometimes called human-centered machine learning designers -- are responsible for using data to train ML models effectively. They work with data scientists to acquire and validate required data sets, deliver data to ML models, test ML models, and provide feedback or optimizations to the models' decision-making. ML trainers are experts in ML behavior, understand how the models operate, and help predict the models' output -- also known as explainability -- to help the business maintain adequate AI performance and AI governance.
ML designers and trainers have strong programming knowledge in languages such as Python for model building and SQL for database management. These roles require a substantial ML technology background supported by an excellent understanding of data sources and quality, as well as skills in system and software design.
Interface specialists
Interface specialist is a broad term referring to a wide range of jobs intended to support complex ML models. For example, computational linguists can help ML models better understand human speech and its nuances, as well as imagine ways to make ML model speech features more intuitive and helpful to users. Similarly, computer vision specialists can help improve object detection and recognition for tasks such as facial or object recognition. Interface specialists also work on ways for humans to interact with ML and AI through text, spoken and visual context prompts.
Speech, language and visual interactions are complex and highly nuanced. Consequently, interface specialist roles generally require a strong mastery of ML and deep learning technologies, along with strong mathematics and statistics skills. Programming expertise can be helpful but secondary.
Specific requirements can vary with the scope and goals of the project. For example, an ML/AI linguist role might also demand comprehensive knowledge of natural language processing and LLMs along with fluency in multiple human languages for projects such as translation and interpretation. Similarly, a computer vision role might require additional skills in image composition, lighting, and other image and video processing elements.
Software developers
Software developers are vital to any ML or AI project. Developers are mainly responsible for building, testing and maintaining the array of code related to ML models, libraries, interfaces, APIs and other software elements involved in the complete ML/AI platform. Developers might work directly on ML models as directed by ML engineers, but might also work on code indirectly related to the ML project.
Developers require strong programming and testing skills in numerous major languages, as well as additional skills with data structures, some knowledge of ML models and algorithm design, and mathematics for data analytics such as statistics and probability.
Business specialists
Business specialists are ML users tasked with using one or more ML platforms to execute data analytics for the benefit of the business. For example, business specialists might run financial projections, research market opportunities, optimize business workflows or audit business operations. Business specialists are often experts in prompt engineering and work closely with data scientists and ML trainers.
Business specialists typically focus on business skills and education, relying on ML and related skills as a means of benefiting the business. Their feedback and use cases can help to drive further ML/AI development. ML/AI business specialists should possess some light programming knowledge to aid in prompt engineering, along with some data modeling and visualization expertise.
ML education and learning opportunities
ML is a demanding pursuit that involves several complex disciplines. Consequently, there is no single degree or course that will forge a successful ML expert. Rather, ML careers flourish through a solid educational foundation, followed by years of job training and further study. However, there are several educational options that can help to jumpstart an ML career and fine-tune long-term goals.
Start with a traditional BS degree
Most major colleges and universities around the world offer a traditional undergraduate bachelor of science degree in core disciplines such as software engineering, computer science or information technology. These degrees provide common onramps for careers related to programming and IT infrastructure. Data science specialists can often start with degree programs in mathematics and data analytics.
Look for ML-specific courses and degrees
Given the incredible growth of ML and AI technologies, many educational institutions today offer courses more specifically tailored to ML and AI knowledge. For example, Kansas State University offers an online bachelor's degree in machine learning and autonomous systems, while St. John's University in Queens, N.Y., offers a bachelor's degree in computing and machine learning.
Look for ML certificates or boot camps
Professionals who already possess a degree can find options for ML-specific certifications or boot camps intended to build ML knowledge and skills much faster than a traditional degree. For example, the Executive Education program at the University of California, Berkeley, offers a six-month professional certificate in ML and AI, while the University of Texas at Austin offers a six-month program in AI and ML. Similar options exist for many colleges and universities worldwide.
Look for shorter certifications for specializations
Certificates can be excellent options for busy professionals with specific ML educational needs and career goals. Professionals seeking to develop more specific ML- and AI-related skills can find varied certificate programs for continuing education and professional development. For example, organizations like Coursera offer course series, including a Deep Learning Specialization and Machine Learning Specialization, among other detailed individual courses. Similarly, providers like Udemy offer courses including Machine Learning and Deep Learning in Python and R, Machine Learning A-Z, and many others.
Get started with machine learning
There is no single roadmap to a successful ML career. The technologies are so new and constantly evolving that no career path is suitably codified, and ML project types and scope can vary dramatically by business and industry vertical. However, there are a few guidelines that can help technical professionals guide their career into a machine learning specialization.
- Start with a degree. Most ML-related jobs involve programming, mathematics and data science, so it can help to start with a traditional undergraduate and advanced degree path. Although formal degrees carry varying importance to different employers, possessing a traditional degree is one of the few standard benchmarks of entry-level proficiency in the rapidly evolving ML field.
- Get into an entry-level role. ML careers are built over the course of years and involve considerable hands-on experience. ML practitioners almost always start their careers with a foundational role, whether in programming or data science. Look for ML-related opportunities and support ML teams working on business projects. This is where practitioners can receive mentoring and learning opportunities in their related disciplines.
- Build ML education. Take advantage of new ML learning opportunities whenever you can. This might involve mentoring and educational presentations within your organization's ML group, as well as outside certificate and boot camp coursework to advance specific ML knowledge and build new techniques. For example, explore a side project building an ML model in a new programming language or experiment with new formatting and syntax tricks for prompt engineering.
- Translate knowledge into practice. As ML opportunities and responsibilities increase over time, put new education and techniques into practice in actual ML projects. Share new ideas with colleagues; create new algorithmic approaches for an ML model; find ways to improve data quality or compute processing efficiency; help colleagues solve ML problems and overcome challenges; and so on. Practice not only reinforces ML learning but also raises professional visibility as an active ML contributor.
- Look for new ML roles. Over time, you'll establish a broad array of ML knowledge matched with a growing list of ML project involvement. Document the successes, study and learn from the failures, and use that flourishing resume to pursue new and challenging ML roles within the business -- you might even be ready to move on to other ML job opportunities.
Stephen J. Bigelow, senior technology editor at TechTarget, has more than 20 years of technical writing experience in the PC and technology industry.