Claude with Google Cloud Vertex AI by Anthropic
7h 26mIntermediate2026-04-02
Authors

Anthropic
Course details
This course provides comprehensive technical training on integrating and deploying Claude AI models through Google Cloud Vertex AI. Developers will learn to implement Claude API capabilities, from basic request handling to advanced features including tool use, retrieval-augmented generation (RAG), and the Model Context Protocol (MCP). The curriculum covers practical implementation patterns, performance optimization techniques, and production-ready workflows for building AI-powered applications.
Prerequisites:
Proficiency in Python programming
Basic knowledge of handling JSON data
A Google Cloud account with Vertex AI access
Learning objectives
Set up and configure Claude models through Google Cloud Vertex AI
Implement multi-turn conversations with proper message handling and context management.
Design and evaluate prompts using systematic testing workflows and automated grading techniques.
Apply prompt engineering principles including XML tag structuring, example-based learning, and output control.
Build tool-use implementations enabling Claude to interact with external functions and APIs.
Develop RAG pipelines using text chunking, embeddings, BM25 search, and contextual retrieval techniques
Use advanced Claude features including vision capabilities, PDF processing, citation generation, and prompt caching.
Implement Model Context Protocol (MCP) for creating custom tools, resources, and prompt templates.
Configure and deploy Anthropic Apps including Claude Code for automated development tasks.
Design agent-based workflows with parallelization, chaining, and routing patterns for complex AI systems.
Prerequisites:
Proficiency in Python programming
Basic knowledge of handling JSON data
A Google Cloud account with Vertex AI access
Learning objectives
Set up and configure Claude models through Google Cloud Vertex AI
Implement multi-turn conversations with proper message handling and context management.
Design and evaluate prompts using systematic testing workflows and automated grading techniques.
Apply prompt engineering principles including XML tag structuring, example-based learning, and output control.
Build tool-use implementations enabling Claude to interact with external functions and APIs.
Develop RAG pipelines using text chunking, embeddings, BM25 search, and contextual retrieval techniques
Use advanced Claude features including vision capabilities, PDF processing, citation generation, and prompt caching.
Implement Model Context Protocol (MCP) for creating custom tools, resources, and prompt templates.
Configure and deploy Anthropic Apps including Claude Code for automated development tasks.
Design agent-based workflows with parallelization, chaining, and routing patterns for complex AI systems.
Concepts
Introduction
- Welcome to the course
Anthropic Overview
- Overview of Claude models
Accessing Claude with the API
- Accessing the API
- Making a request
- Multi-turn conversations
- Chat exercise
- System prompts
- System prompts exercise
- Temperature
- Response streaming
- Controlling model output
- Structured data
- Structured data exercise
Prompt Evaluation
- Prompt evaluation
- A typical eval workflow
- Generating test datasets
- Running the eval
- Model-based grading
- Code-based grading
- Exercise on prompt evals
Prompt Engineering Techniques
- Prompt engineering
- Being clear and direct
- Being specific
- Structure with XML tags
- Providing examples
- Exercise on prompting
Tool Use with Claude
- Introducing tool use
- Project overview
- Tool functions
- Tool schemas
- Handling message blocks
- Sending tool results
- Multi-turn conversations with tools
- Implementing multiple turns
- Using multiple tools
- The batch tool
- Tools for structured data
- The text edit tool
- The web search tool
Retrieval-Augmented Generation
- Introducing retrieval-augmented generation (RAG)
- Text chunking strategies
- Text embeddings
- The full RAG flow
- Implementing the RAG flow
- BM25 lexical search
- A multi-index RAG pipeline
- Reranking results
- Contextual retrieval
Features of Claude
- Extended thinking
- Image support
- PDF support
- Citations
- Prompt caching
- Rules of prompt caching
- Prompt caching in action
Model Context Protocol (MCP)
- Introducing MCP
- MCP clients
- Project setup
- Defining tools with MCP
- The server inspector
- Prompts in the client
- MCP review
- Implementing a client
- Defining resources
- Accessing resources
- Defining prompts
Anthropic Apps - Claude Code and Computer Use
- Anthropic apps
- Claude Code setup
- Claude Code in action
- Enhancements with MCP servers
- Parallelizing Claude Code
- Automated debugging
- Computer Use
- How Computer Use works
Agents and Workflows
- Agents and workflows
- Parallelization workflows
- Chaining workflows
- Routing workflows
- Agents and tools
- Environment inspection
- Workflows vs agents
Wrapping Up
- Course wrap-up