AI is one of the most sought-after skill areas in the job market, but the path in can seem confusing. You don’t need to be a genius or have a PhD for most roles. You do need solid foundations, evidence of practical skills and persistence. Here is a realistic roadmap.

Step 1 — Choose a direction

Read about the different AI job roles and pick a target: for example AI engineer, ML engineer or data scientist. You can change later, but a target helps you prioritize.

Step 2 — Build foundations

Programming: Python is essential. Learn data structures, functions, classes, working with files and APIs, and tools like Git.

Math (as needed):

  • linear algebra (vectors, matrices);
  • probability and statistics;
  • basic calculus (derivatives, gradients).

You don’t need to master all the math before starting; learn it alongside practical work, deepening as needed.

Data skills: SQL, pandas and data visualization.

Step 3 — Learn machine learning

Good free and low-cost resources include:

  • fast.ai’s practical deep learning courses, which start with working code;
  • Stanford’s CS231n and other university courses with public materials;
  • Hugging Face courses on transformers, language models and agents;
  • introductory machine learning courses on online learning platforms.

Learn the concepts — training and test data, overfitting, evaluation metrics — and implement simple models yourself.

Step 4 — Build projects

Projects matter more than certificates. Good projects:

  • solve a real problem you care about;
  • use real, messy data;
  • include clear evaluation;
  • are deployed or demoable (a simple web app);
  • are documented with a readme explaining decisions and results.

Examples: a retrieval-augmented assistant over a specific document collection, a model that predicts something from public data with careful evaluation, or a tool that automates a task in a domain you know.

Step 5 — Contribute to open source

Fixing documentation, adding tests or solving small issues in open-source AI projects teaches you real-world code and collaboration, and gives visible evidence of your skills.

Step 6 — Build a portfolio and presence

  • A GitHub profile with a few strong, well-documented projects.
  • Short write-ups or blog posts explaining what you built and learned.
  • A concise CV highlighting projects and measurable results.

Step 7 — Network

Join communities, attend meetups, participate in competitions and hackathons, and talk to people working in roles you want. Many jobs are found through referrals.

Step 8 — Apply strategically

  • Consider adjacent roles — data analyst, software engineer, solutions engineer — as stepping stones.
  • Use your domain expertise: a nurse, lawyer or accountant who learns AI has a rare combination.
  • Prepare for interviews with coding practice, ML concepts and project discussions.

For career changers

Your previous experience is an asset. Look for AI projects in your current job, propose automation or analytics initiatives, and position yourself as the person who bridges your domain and AI.

A realistic timeline

With consistent part-time study, many people build job-ready skills for entry-level roles in roughly one to two years, depending on background and target role.

A sample first-year plan

  • Months 1–3: Python, Git, SQL and data analysis with pandas; one small data project.
  • Months 4–6: core machine learning concepts and a practical deep learning course; two projects with careful evaluation.
  • Months 7–9: specialize — for example language models and retrieval for AI engineering; build and deploy an application.
  • Months 10–12: contribute to open source, polish your portfolio, network actively and start applying.

Adjust the pace to your circumstances; consistency matters more than intensity.

What to watch next

  • Changes in entry-level hiring as AI tools automate parts of coding.
  • Demand for domain experts with AI skills.
  • New certifications tied to AI regulation and governance.
  • Apprenticeships and alternative pathways into AI roles.

Key takeaways

  • Choose a target role to guide your learning.
  • Python, data skills and practical math are the foundations.
  • Free resources like fast.ai, CS231n and Hugging Face courses are excellent.
  • Projects and open-source contributions matter more than certificates.
  • Domain expertise plus AI skills is a strong advantage for career changers.

Frequently asked questions

Am I too old to start in AI?

No. Many people switch careers into AI-related roles, often bringing valuable domain knowledge.

Are bootcamps worth it?

Some are, but quality varies. Check outcomes, curriculum and whether you could learn the same with free resources and discipline.

How much math do I really need?

For applied roles, a working understanding of linear algebra, probability and statistics is usually enough; research needs much more.