662 free lessons · no sign-up needed

Data science & AI,
taught from scratch.

The complete path from your first line of Python to production AI — become a data analyst, data scientist, or AI engineer.

Start with Python — free Browse all 20 courses

Python SQL Machine Learning Deep Learning Generative AI Agentic AI GATE DA

662

lessons

650+

interview Q&A

20

courses

01 Where do you want to land?

Pick the job. We’ll sequence the courses.

The right courses in the right order, taught from scratch. Three headline tracks here; 9 in all.

Best first job in data

Data Analyst

Answer real business questions with data — SQL, pandas, charts that tell the story, and the analytics sense to turn numbers into decisions.

SQL Python Pandas Storytelling with Visualisation Business Analytics

114 lessons · 8–12 weeks See the path

The classic path

Data Scientist

From SQL to statistical rigor to communicating results that change decisions.

Python Pandas Storytelling with Visualisation SQL Math for ML Machine Learning

163 lessons · 12–16 weeks See the path

The 2026 frontier

AI Engineer

Build AI products from first principles — neural nets, transformers, LLMs, RAG and agents, all the way to shipping them in production.

Python Machine Learning Deep Learning NLP & Transformers Generative AI Agentic AI MLOps

308 lessons · 16–22 weeks See the path

See all 9 career paths

02 The full curriculum

20 courses, 662 lessons. All of it free.

In learning order, grouped the way the field actually fits together. Each card opens its first lesson straight away.

Foundation

Python From syntax to production AI apps — the language for everything. 41 SQL Analytics-grade SQL, warehouse dialects, and dbt. 27 Data Structures & Algorithms The algorithmic core every data & AI engineer needs — in Python. 32 Git Version control the way every team actually uses it — branch, merge, recover. 15 Command Line The terminal skills that make every other tool faster. 14

Data & Stats

NumPy The numeric foundation of the entire data stack. 14 Pandas The de-facto tool for tabular data in Python. 13 Business Analytics Turn data into decisions — the analytics an MBA pays for. 21 Math for ML Linear algebra, calculus, and probability with code. 37

EDA & Viz

Storytelling with Visualisation Matplotlib, Seaborn — and how to turn a chart into a story that lands. 12

ML Fundamentals

Machine Learning Trees beat neural nets on most tabular problems — and other truths. 33

Advanced ML

Time Series Forecast what comes next — ARIMA, SARIMA, VAR, Prophet, done right. 14 Recommender Systems How Netflix, Spotify & Amazon decide what you see next. 11

Deep Learning

Deep Learning Neural networks from scratch — the core, no fluff. 28

NLP & Transformers

NLP & Transformers From text features to the transformer — language modeling end to end. 44

MLOps & Production

MLOps Ship models, not notebooks. 28 PySpark Big data at scale — Hadoop → Spark internals → production. 22

AI

Generative AI LLMs in practice — RAG, evals, fine-tuning, self-hosting. 63 Agentic AI Build agents on the LLM — patterns, tools, MCP, multi-agent, production. 71

Exam Prep

GATE DA Crack GATE Data Science & AI — every topic, ground-up to exam level. 122

03 Start with a classic

The lessons readers open first.

Today’s pick Top-N per group

Python 10 min Functions Read the lesson SQL 10 min Window functions Read the lesson Machine Learning 7 min Bias–variance & learning curves Read the lesson Deep Learning 9 min Backprop by hand Read the lesson Pandas 10 min GroupBy Read the lesson Machine Learning 8 min Linear regression Read the lesson

GATE Data Science & AI paper

Preparing for GATE DA? The whole syllabus is here.

122 exam-grade lessons covering every section of the paper — probability, linear algebra, ML, databases, and general aptitude — with solved problems in every lesson.

Start preparing

04 Land the job

Prep & reference, when it counts.

Interview rounds, a quick definition, the syntax you forgot — the fast lane next to the courses.

650+ questions · role-filtered

Interview Q&A

Real questions across SQL, ML, Python, statistics & AI — each with a worked answer and the trap to avoid. Filter by the role you're chasing.

SQLMachine LearningPythonStatisticsDeep LearningNLP & LLMs

Start practising

Glossary

245 data & AI terms in plain English — from overfitting to RAG — each linked to a lesson.

Cheat sheets

6 dense one-pagers for Python, SQL, Pandas, NumPy, Git & scikit-learn — the syntax you actually reach for.

05 From the blog

Activation checkpointing makes GPU memory a scheduling decision

Large-model training is constrained by more than weights. Activation checkpointing changes which forward tensors survive, trading recomputation for a smaller peak-memory footprint.

RAG poisoning is an evidence-integrity problem

A retrieval system can reason perfectly from corrupted evidence. Defending RAG means governing what enters the corpus, preserving provenance, isolating tenants, and treating retrieved text as untrusted data.

Infrastructure · Jun 18

Six system-design boundaries that prevent category mistakes

Stateless vs stateful, Lambda vs ECS, database vs cache, queue vs stream, retrieval vs reranking, and monitoring vs tracing—explained as operational contracts.

Agents · Jun 18

Token theft in AI agents is an architecture failure

Prompt injection gets the attention, but credentials turn a confused model into an authenticated attacker. The fix is to keep tokens out of model context and authorize every action at runtime.

From scratch, with the math

Every idea is built up from first principles — derivations traced step by step, then turned into working code you can follow line by line.

Every output is real

Code samples ship with their actual verified output — run before publishing, never guessed. What you see is what the code does.

One clear idea at a time

Short lessons in a deliberate order — each one hands off to the next, so the whole field clicks together instead of piling up.

Read more about datarekha · questions? get in touch