Phase 01
Python + Data
Prepare the data
Goal
Become fluent in the tools used to inspect and transform data.
0%
Progress0%
0%
Start with Phase 01.
Phase 02
Math + Statistics
Reason about uncertainty
Goal
Develop the mathematical intuition required to understand model behavior.
0%
Phase 03
Classical ML
Learn the core algorithms
Goal
Understand how and when to use the major supervised and unsupervised approaches.
0%
Phase 04
Evaluation + Features
Make models trustworthy
Goal
Build robust validation and meaningful features.
0%
Phase 05
Deep Learning
Modern neural models
Goal
Move beyond tabular ML when the problem demands it.
0%
Phase 06
Production ML
Operate the system
Goal
Ship models as reliable software.
0%
Phase 07
Portfolio + Interviews
Prove your skills
Goal
Turn ML knowledge into evidence employers can evaluate.
0%
How to use it
Learn → build → prove.
01
Follow the phases in order. Open a checkpoint when you're ready to dive deeper.
02
Use the build checkpoints to turn concepts into working projects.
03
Mark progress locally and use the final phase to turn learning into evidence.
Roadmap notes Read the reasoning behind the sequence
How to use this roadmap
Start with Python and data work, then build mathematical intuition, classical ML, evaluation skills, deep learning, and production ML. Use the visual checkpoints as the main interface.
What this path optimizes for
Correct experimentation, reliable evaluation, production thinking, and the ability to explain model trade-offs.
Progress rule
Prioritize the core checkpoints. Pick a specialty only after you can complete an end-to-end ML workflow.
Checkpoint