Updated July 2026

Data Scientist Roadmap

From Python, SQL, and statistics to A/B testing, applied ML, and business communication — a structured, free path to becoming a data scientist.

8 learning stages
10+ free courses & tutorials
7–10 mo part-time pace
01 —

Python, SQL & Statistics Foundations

4–6 weeks

The three languages every data scientist needs to be fluent in: Python, SQL, and statistics.

What you'll learn

  • Python for data science (Pandas, NumPy)
  • SQL: joins, window functions, aggregations
  • Descriptive statistics and probability
  • Linear algebra and calculus essentials

Courses

Machine Learning Foundations — Free · Start here
02 —

Data Collection, Cleaning & EDA

4–6 weeks

Real data is never clean. This is where most of a data scientist's actual time goes.

What you'll learn

  • Data collection: APIs, web scraping, and database queries
  • Handling missing data, outliers, and duplicates
  • Exploratory data analysis (EDA)
  • Data quality validation and profiling

Courses

Data Science Specialization — Free
03 —

Statistical Analysis & Hypothesis Testing

4–6 weeks

The rigor that separates a data scientist from someone who just runs `.describe()`.

What you'll learn

  • Hypothesis testing, p-values, and confidence intervals
  • Regression analysis and interpreting coefficients
  • Bayesian thinking and Bayesian networks
  • Common statistical pitfalls (multiple comparisons, Simpson's paradox)
04 —

Data Visualization & Storytelling

2–4 weeks

An analysis nobody understands doesn't drive a decision. This stage is about being understood.

What you'll learn

  • Choosing the right chart for the question being asked
  • Building dashboards stakeholders actually use
  • Communicating uncertainty honestly
  • Avoiding misleading visualizations
05 —

Experimentation & A/B Testing

3–4 weeks

The skill that turns a data scientist into a decision-maker, not just a reporter.

What you'll learn

  • Designing a valid experiment and choosing a metric
  • Power analysis and sample size calculation
  • Common pitfalls: peeking, novelty effects, multiple comparisons
  • Reading a result and making a clear recommendation

Free Tutorials

A/B Testing →
06 —

Applied Machine Learning for Data Scientists

6–8 weeks

Enough ML to build, evaluate, and explain a model — without needing to become an ML engineer.

What you'll learn

  • Regression and classification for business problems
  • Feature engineering from raw business data
  • Time-series analysis and forecasting
  • Model evaluation metrics stakeholders can understand

Courses

Intermediate Machine Learning — Free
07 —

Business Intelligence, Communication & Portfolio

3–4 weeks

Ship the artifacts that get you hired: a dashboard, a memo, and a portfolio.

What you'll learn

  • Building a BI dashboard end to end
  • Translating analysis into a recommendation memo
  • Presenting findings to non-technical executives
  • Building a portfolio that demonstrates real impact
08 —

Certification & Interview Preparation

Ongoing

Prove it end to end, get certified, and prepare for every category of question the interview loop will throw at you.

What you'll learn

  • SQL and statistics interview questions
  • Case-study and product-sense questions
  • A/B testing and experimentation questions
  • Communicating a case study to a non-technical panel

Courses

Data Science Specialization Certification — Free · Get certified

Ready to start?

All courses are free and backed by real open source products you can use immediately.

Start: Data Science Specialization Get Certified → Interview Prep →
AI Engineer Path → ML Engineer Path → AI Research Scientist Path → MLOps Engineer Path → AI Product Manager Path → Forward Deploy Engineer Path →