beginner → Portfolio-ready
A structured AI and machine-learning journey from Python, data, and mathematical foundations through classical machine learning, deep learning, responsible evaluation, MLOps, deployment, and a portfolio-ready capstone.
Use this overview to understand the sequence. Open each stage below for course details.
Included published courses are activated together. Your course certificates remain individual, and completing every required certified course unlocks the roadmap distinction.
Required courses build the core skill. Optional courses let you specialise.
Build the programming, data, probability, statistics, and linear-algebra foundation used throughout machine learning.
Learn practical Python, NumPy, pandas, notebooks, and reliable data-processing habits.
Develop the mathematical intuition needed to understand data, optimisation, uncertainty, and model behaviour.
Prepare real datasets and solve supervised and unsupervised learning problems with reproducible workflows.
This sample mirrors the issued certificate. Your name, verified results, QR code, and verification ID appear only after every required course certificate is earned.
Clean, transform, visualise, and investigate real datasets before modelling.
Train, tune, and compare regression, classification, clustering, and dimensionality-reduction models.
Develop neural-network intuition while learning to evaluate, explain, and improve models responsibly.
Build neural networks and apply modern deep-learning workflows to practical problems.
Measure model quality, inspect failure modes, explain predictions, and account for fairness, safety, and privacy.
Serve, monitor, and present a complete machine-learning product built for real-world use.
Package models behind APIs and create repeatable training, deployment, monitoring, and versioning workflows.
Define a problem, build and evaluate a model, deploy the solution, and communicate its limitations and impact.
Awarded for meeting the published learning and assessment requirements. Preview only — it has no verification ID and cannot be used as a credential.