Choose your certification and start revising.

Understand tokens, temperature, top-k and top-p sampling, and how language models generate text.

Map foundation models to business tasks: text generation, summarization, classification, code, and multimodal.

Know the difference between pre-training, fine-tuning, prompt engineering, and RAG — and when each fits a scenario.

Gemini is Google's multimodal foundation model family; Vertex AI is the unified ML platform for building, deploying, and managing models.

Distinguish Vertex AI Studio for prompt development, Model Garden for accessing Google and partner models, and Agent Builder for workflow orchestration.

Google Workspace AI automates productivity tasks across Docs, Sheets, Meet, and Gmail using generative AI.

Google's AI principles cover fairness, interpretability, privacy, security, and societal benefit — know them by purpose, not memorized list.

AI governance includes explainability, bias detection, human oversight, and audit trails.

Avoid overengineering answers: this is a foundational certification focused on strategy and adoption, not deep model architecture.