Data science pays well and the best resources are free. The roadmap below takes you from zero to portfolio-ready in 6–12 months, using only free tools and courses. No degree required.
The learning roadmap
Data science has three layers: coding (Python and SQL), statistics (the math behind predictions), and tools (visualization and libraries). Learn them in order — each builds on the last.
1. Learn Python (months 1–2)
- freeCodeCamp's Python course — free, project-based, covers fundamentals.
- Harvard CS50P — free on edX, teaches Python from scratch.
- Kaggle Learn — short, practical Python tutorials for data work.
Focus on: variables, loops, functions, lists, dictionaries, and file reading. You do not need to master everything — learn enough to work with data.
2. Learn SQL (month 2–3)
Every data role requires SQL. Practice on SQLZoo (free), HackerRank SQL (free), and Mode Analytics SQL tutorial (free). Learn SELECT, JOIN, GROUP BY, window functions, and subqueries.
3. Statistics fundamentals (months 3–4)
- Khan Academy Statistics — free, visual, covers distributions, hypothesis testing, regression.
- StatQuest YouTube — free, explains complex stats in plain English.
Key concepts: mean/median/mode, standard deviation, correlation vs causation, p-values, confidence intervals, basic regression.
4. Data manipulation with pandas (months 4–5)
pandas is the core Python library for data work. Learn to load CSVs, clean data, filter rows, group by categories, and merge datasets. Kaggle's pandas course is free and practical.
5. Data visualization (months 5–6)
- Matplotlib and Seaborn — Python libraries for charts and plots.
- Tableau Public — free version for interactive dashboards.
- Google Looker Studio — free, creates shareable dashboards from spreadsheets.
6. Intro to machine learning (months 6–8)
Kaggle's free "Intro to Machine Learning" course covers decision trees, random forests, and model evaluation. Practice on Kaggle competitions — real datasets, real problems, free GPU.
7. Build a portfolio (months 8–12)
Projects prove skill. Build three to five projects and publish them on GitHub:
- Analyze a real dataset (public data from data.gov, Kaggle, or WHO).
- Build a dashboard that tells a story.
- Solve a Kaggle competition end-to-end.
- Create a web app that lets users explore data (Streamlit is free).
8. Get hired or freelance
Data roles: junior data analyst, business intelligence analyst, data scientist. Freelance platforms (Upwork, Toptal) pay $30–100/hr for data work. Start with small projects — cleaning datasets, building dashboards, automating reports.
FAQ
Do I need a degree for data science?
No. Many data scientists are self-taught. A strong GitHub portfolio with real projects speaks louder than a degree. Companies like Google, Apple, and IBM have dropped degree requirements.
How long does it take to learn data science?
With 1–2 hours of daily practice, most people reach job-ready level in 6–12 months. The key is consistency and building projects, not just watching courses.
What laptop do I need?
Any laptop that runs a browser. Google Colab provides free GPU and Python in the cloud — no installation needed. For local work, 8GB RAM is enough to start.
