What do students learn in the SAAI Research Program?

Students progress through a connected research journey rather than a collection of disconnected AI tools. They first learn how text and data are represented through concepts such as tokenisation, vectors and contextual meaning. They then study datasets, labels, features, training and testing splits, model selection, predictions, confusion matrices and performance metrics. As they advance, students design controlled experiments, compare alternative models, identify limitations, record observations and explain why one approach performs differently from another. Research skills include defining a question, forming a testable hypothesis, documenting methodology, organising evidence, interpreting results and avoiding unsupported claims. The final stage requires the learner to convert technical work into a coherent research artifact such as a notebook, report, paper, model comparison or working prototype, followed by a defence of the choices made.