Application Development Training with Claude

About Education

This training course is designed for developers, product managers, and tech enthusiasts who want to bring a new dimension to their application development processes using Anthropic’s powerful AI model, Claude. Throughout the course, we will cover Claude’s API capabilities, natural language processing (NLP), data analysis, content generation, automation, and AI-powered application scenarios in detail.

Participants will gain hands-on knowledge on topics such as how to develop interactive applications with Claude, how to process user input, and how to interpret Claude’s outputs.

 

Prerequisites

• Basic knowledge of software development (Python, JavaScript or similar language preferred)
• Basic information about API usage
• General familiarity with artificial intelligence and machine learning concepts (preferred but not required)

 

Duration of Training

• Instructor-led: 5 days of hands-on training
• On-site

 

Who Should Participate?

• Application developers and software engineers
• Product managers and digital transformation professionals
• Entrepreneurs who want to integrate AI solutions into their products

 

Educational Content

Day 1 – Fundamentals and Getting Started

Morning – Claude Introduction And an Overview of AI-Powered Development Tools

• Claude model ailesine genel bakış (Opus, Sonnet, Haiku serileri)
• Choosing the right model: balancing performance and cost
• Map of AI-powered development tools: When to use Claude API, Claude Code, and Claude Desktop

Before Lunch – API Basics

• Obtaining an API key and authorisation
• Messages API structure and message flow
• Practical: Sample application with Python/Node

Afternoon – Introduction to Prompt Engineering

• The Basic Principles of Writing Effective Prompts
• Examples of information extraction and structuring
• Application: prompt design that produces structured output

 

Day 2 – Advanced API Usage and Multimodality

Morning – Performance and Scalability

• Message Batching (Message Batches API)
• Prompt caching and performance
• Code example: adding a caching layer

Before Lunch – File, PDF, and Image Analysis

• File upload and analysis
• PDF content processing
• Visual data processing, graph/chart interpretation

Afternoon – Structured Output and Validation

• Few-shot prompt design
• Tool use and the production of structured output
• Validation and retry loops

 

Day 3 – Vehicle Design, Introduction to MCP and Agents

Morning – Fundamentals of Tool Design

• Claude's driving characteristics
• Best practices in vehicle design: clear explanations, understandable error messages
• Selection of built-in tools

Before Lunch – Model Context Protocol (MCP)

• What is MCP and what does it do?
• Use with backend integration
• Connecting MCP servers to the project

Afternoon – Introduction to Agents

• Agentic loop: How Claude repeats the steps of planning a task, invoking an agent, and evaluating the result
• Understanding when the cycle stops: when the model signals “job complete” or “I need a tool”
• Developing a JSON-based task agent
• Mini agent application: resume parser

 

Day 4 – Developing with Claude Code

Morning – The Basics of Claude Code

• What is Claude Code, and how does it differ from using an API?
• Defining project context with the CLAUDE.md file
• Installation and Basic Commands

Before Lunch – Customization and Work Modes

• Special slash commands and skills
• The difference between plan mode and direct execution
• Defining Scope with Path-Specific Rules

Afternoon – Real-World Project Implementation

• Iterative improvement techniques
• Overview of CI/CD Workflows
• Application: Completing a small project from start to finish using Claude Code

 

Day 5 – Multi-Step Agent Systems, Context Management, and RAG

Morning – Multi-Step Agent Systems

• Smart agents with multi-step planning
• The logic of task decomposition
• Collaboration Among Multiple Agents: Basic Concepts

Before Lunch – Context Management and Reliability

• Effectively managing the context of a conversation
• When can the output be trusted, and when should it be subject to human review?
• Escalation and referral to the appropriate resource in uncertain situations

Afternoon – RAG and the Closing Project

• Using Claude with External Information Sources (RAG)
• Setting Up a Q&A System with Database/Document Integration
• Closing Project: End-to-End Claude-Powered Workflow and Presentation

What You Will Gain at the End of Training

• Use the Claude API effectively
• Multimodal and tool-agent systems
• Advanced prompt design
• Artificial intelligence evaluation and production readiness