Google Cloud AI Course

Lesson 1 • Core GCP Services Overview

Surveys Compute Engine, Cloud Storage, BigQuery, and networking primitives. Provides context for where AI services fit within the broader platform.

Lesson 2 • Cloud Computing Core Concepts

Covers IaaS, PaaS, and SaaS models alongside GCP's global infrastructure. Establishes the mental model needed for all subsequent cloud AI work.

Lesson 3 • GCP Console and CLI Setup

Hands-on orientation to the GCP Console, Cloud Shell, and gcloud CLI. Enables students to provision and manage resources from day one.

Lesson 1 • BigQuery for AI Workloads

Teaches SQL-based data exploration, partitioning, and feature extraction in BigQuery. Prepares structured datasets for Vertex AI training jobs.

Lesson 2 • Data Ingestion Strategies on GCP

Covers batch and streaming ingestion using Cloud Storage, Pub/Sub, and Dataflow. Connects raw data sources to downstream AI workflows.

Lesson 3 • Dataset Versioning and Governance

Introduces Vertex AI Datasets, data lineage, and access controls for reproducible AI. Ensures compliance with data quality and auditability requirements.

Lesson 1 • Natural Language Processing APIs

Applies Cloud Natural Language API for entity recognition, sentiment, and syntax analysis. Connects text analytics to real business classification tasks.

Lesson 2 • Speech and Translation Services

Integrates Speech-to-Text, Text-to-Speech, and Cloud Translation APIs into pipelines. Enables multilingual and voice-driven application features.

Lesson 3 • API Integration and Error Handling

Covers authentication patterns, quota management, retry logic, and SDK usage. Ensures production-grade reliability when calling GCP AI APIs.

Lesson 4 • Video Intelligence API

Extracts labels, shot changes, and explicit content signals from video using the Video Intelligence API. Extends AI capabilities to multimedia content pipelines.

Lesson 1 • AutoML Image Classification

Builds and evaluates an image classification model using Vertex AI AutoML Vision. Covers dataset preparation, training budget, and evaluation metrics.

Lesson 2 • AutoML Text and NLP Models

Creates custom text classification and entity extraction models with AutoML Natural Language. Applies trained models to domain-specific document processing.

Lesson 3 • AutoML Concepts and Use Cases

Explains transfer learning, neural architecture search, and when AutoML outperforms manual ML. Frames AutoML as a strategic tool within the AI tier selection.

Lesson 4 • Deploying and Monitoring AutoML Models

Deploys AutoML models to Vertex AI Endpoints and monitors prediction quality over time. Closes the loop from training to production-ready inference.

Lesson 1 • Hyperparameter Tuning with Vertex AI

Automates hyperparameter search using Vertex AI Vizier and custom tuning jobs. Reduces manual experimentation and improves model performance systematically.

Lesson 2 • Model Registry and Versioning

Registers trained models in Vertex AI Model Registry with version labels and metadata. Provides a governed catalog for promotion to staging and production.

Lesson 3 • Vertex AI Workbench and Notebooks

Sets up managed Jupyter notebooks in Vertex AI Workbench for iterative ML development. Establishes the interactive environment for all custom training work.

Lesson 4 • Training with TensorFlow and PyTorch

Packages TensorFlow and PyTorch training scripts for Vertex AI custom training jobs. Covers containerization, hardware selection, and distributed strategies.

Lesson 1 • Model Monitoring in Production

Configures Vertex AI Model Monitoring for skew, drift, and feature attribution alerts. Closes the observability loop for deployed models in live traffic.

Lesson 2 • Vertex AI Pipelines Execution

Submits, monitors, and debugs pipeline runs on Vertex AI managed infrastructure. Connects pipeline outputs to the Model Registry and Endpoints.

Lesson 3 • Building Pipelines with Kubeflow

Authors reusable pipeline components using the Kubeflow Pipelines SDK and KFP v2. Translates notebook experiments into production-grade DAG workflows.

Lesson 4 • MLOps Principles and Maturity Levels

Defines MLOps maturity levels from manual to fully automated CI/CD/CT pipelines. Provides a framework for assessing and improving organizational ML practices.

Lesson 1 • Grounding and Retrieval-Augmented Generation

Implements RAG pipelines using Vertex AI Search and Vector Search to ground LLM responses. Reduces hallucination and connects models to enterprise knowledge bases.

Lesson 2 • Generative AI Fundamentals on GCP

Explains large language model architecture, tokenization, and the Gemini model family. Grounds generative AI concepts in GCP's specific offerings and APIs.

Lesson 3 • Prompt Engineering Techniques

Applies zero-shot, few-shot, chain-of-thought, and system instruction prompting strategies. Directly improves output quality for downstream application development.

Lesson 4 • Building Generative AI Applications

Constructs end-to-end generative AI apps using the Vertex AI SDK, LangChain, and Cloud Run. Delivers scalable, API-backed generative AI services.

Lesson 1 • Data Privacy and Compliance

Applies Cloud DLP, data residency controls, and anonymization techniques to AI datasets. Aligns AI data handling with privacy regulations and organizational policies.

Lesson 2 • Security Controls for AI Workloads

Configures VPC Service Controls, CMEK, and IAM least-privilege for AI pipelines. Protects sensitive training data and model artifacts from unauthorized access.

Lesson 3 • AI Governance Frameworks

Establishes model governance policies, approval workflows, and audit trails using GCP tools. Enables organizational accountability across the full AI lifecycle.

Lesson 4 • Responsible AI and Bias Mitigation

Applies Google's Responsible AI principles to detect, measure, and mitigate model bias. Ensures AI outputs meet fairness and transparency standards.