Machine Learning
Machine learning is how software learns from data. Instead of writing rules that tell a program what to do, you give it examples and it figures out the patterns. A spam filter doesn't have a list of spam words. It learned what spam looks like from millions of emails.
ML practitioners work across two connected worlds. Classical ML means training models on structured data: predicting which customers will cancel, detecting fraud in transactions, forecasting demand. LLM applications means building systems around large language models: retrieval-augmented generation, evaluation pipelines, prompt engineering, fine-tuning. Most practitioners navigate both, often in the same project.
The track
Projects span classical ML and LLM applications. You'll direct AI to build systems for fictional clients, then verify the output, catch the mistakes, and make the design decisions that determine whether a system succeeds or fails in production.
The skill you're building isn't machine learning from scratch. It's directing AI to build ML systems and verifying the result: knowing what to ask for, what inputs matter, how to evaluate output, and when to push back.
Before you start
- Read the Introduction: what the field is, how the work flows, what tools you'll use
- Complete the Platform Setup: accounts, terminal, Claude Code, Git (same for all tracks)
- Complete the ML Setup: Python, ML libraries, and a hands-on demo