Definition
AI careers fall into a few main roles. Data analysts explain what happened. Data scientists build predictive models. ML engineers run models in production. GenAI engineers build products on LLMs. Research scientists invent new methods. Each needs a different mix of skills.
In simple words
"Working in AI" covers very different jobs, the way "working in a hospital" covers surgeons, nurses and accountants. Choose by what you enjoy day to day.
Why it matters
Choosing a role early saves you from learning everything. This course's paths are built around these roles, and you can change your mind at any time.
Key terms
- Data analyst
- Someone who studies data to answer questions and explain what happened, often with charts and dashboards.
- Data scientist
- Someone who builds models that predict or classify, and runs experiments to test ideas.
- Machine learning engineer
- Someone who turns models into reliable systems that many people can use.
- AI product manager
- Someone who decides what an AI product should do, for whom, and how its success is measured.
- AI ethics and policy specialist
- Someone who works on fairness, safety, privacy and laws for AI.
- Portfolio
- A collection of projects you have built and can show to others.
Explained in detail
Many roles, not one
People often imagine an "AI job" as one person at a computer inventing robots. In reality, building and using AI takes teams with many different skills. Some roles are deeply technical: data engineers who collect and organise data, data scientists who build models, machine learning engineers who make those models run reliably for millions of users, and research scientists who invent new methods. Other roles need different strengths: product managers who decide what to build, designers who make AI tools easy to use, ethics and policy specialists who think about fairness and law, and trainers who teach people to use AI well.
AI in every profession
Perhaps the largest group of AI careers is not "AI jobs" at all. Doctors who use AI to read scans, farmers who use crop-advice apps, lawyers who use AI to search documents, journalists who check deepfakes and teachers who design lessons with AI all need to understand what AI can and cannot do. In these jobs, knowing your own field deeply and knowing how to use AI wisely is a powerful combination. A doctor who understands both medicine and AI's limits is far more valuable than an AI tool or a technologist alone.
Roles and what they need
| Role | Enjoys | Builds on |
|---|---|---|
| Data analyst | Finding answers in numbers and explaining them | Maths, statistics, charts, clear writing |
| Data scientist | Experiments and predictions | Statistics, programming, machine learning |
| Machine learning engineer | Making systems that work reliably at scale | Programming, software engineering |
| AI product manager | Understanding users and leading teams | Communication, problem scoping, basic AI knowledge |
| Designer for AI tools | Making technology easy and fair to use | Design thinking, empathy, user testing |
| Ethics and policy specialist | Questions of fairness, rights and law | Law, social science, philosophy, AI basics |
| Research scientist | Inventing new methods | Advanced maths, deep learning, reading research |
How to prepare, starting now
- Build strong foundations in maths, reading, writing and logical thinking. Every role uses them.
- Stay curious about a field you care about, whether health, farming, music or law. AI is most useful when combined with real knowledge of a problem.
- Make small projects, such as a classifier built with a no-code tool or a data story about your town, and write about what you learned.
- Practise explaining AI to others. Communication matters in every role.
- Keep learning. Tools change quickly; the habit of learning lasts.
You do not need to choose your career now. The later levels of this course follow different paths, and you can try several before deciding. What matters at this stage is understanding the landscape, discovering what you enjoy, and building the habits that every path needs.
How it works
| Role | Day to day | Core skills |
|---|---|---|
| Data Analyst | Dashboards and answering business questions | SQL, Excel, programming, statistics |
| Data Scientist | Experiments and predictive models | Programming, statistics, ML |
| ML Engineer | Deploying and scaling models | Programming, software engineering, MLOps |
| GenAI Engineer | Building RAG and agent products | Programming, APIs, LLMs, evaluation |
| Research Scientist | Reading and writing papers, new methods | Maths, deep learning, PyTorch |
| Data Engineer | Building data pipelines | SQL, Spark, cloud |
Try it yourself
One product, many jobs
A team builds a crop-advice app. Watch the work pass from person to person.
Career cards: what's the daily work?
Choose a career by what you'd enjoy day to day. Match each role to its daily work.
- Data Analyst
- Data Scientist
- ML Engineer
- GenAI Engineer
- Research Scientist
- Data Engineer
Common mistakes
- Collecting certificates with no projects to show.
- Learning only models and skipping SQL and basic engineering, which most jobs need.
Check yourself
Summary
- AI has several distinct roles.
- Choose by the daily work you enjoy.
- Projects you can show count most.
Finished reading?