Get in Touch
 Duration 14 hours

Course Outline

Introduction

  • Distinctions between workflows and agents
  • Architectural overview and common application scenarios

Claude API Essentials

  • Details on the Messages API, available models, and parameters
  • Crafting system prompts and defining structured outputs
  • Implementing streaming and robust error handling
  • Lab: Initial API integration

Tool Integration

  • Specifying tools and their schemas
  • Understanding the tool-use loop and error recovery mechanisms
  • Lab: Creating an agent that interacts with internal company APIs

MCP and Agent Skills

  • Overview of MCP servers and clients
  • Leveraging existing servers and developing a basic custom server
  • Implementing Agent Skills for reusable specialized knowledge
  • Lab: Connecting an agent to a specific data source

Claude Agent SDK

  • Managing the agent loop, sessions, and context
  • Configuring permissions, hooks, and subagents
  • Lab: Developing a research and reporting agent

Cost, Quality, and Safety

  • Optimizing with prompt caching and managing token budgets
  • Implementing guardrails and human-in-the-loop approval steps
  • Conducting evaluations using a small, focused test set
  • Setting up logging and monitoring systems

Transitioning to Production

  • Deployment strategies and secure secrets management
  • Best practices for maintenance and version control
  • Creating a roadmap for participants' specific use cases

Requirements

  • Proficiency in programming with Python or TypeScript
  • Experience with REST API calls and JSON data handling
  • A foundational grasp of large language models

Intended Audience

  • Software developers and data engineers integrating AI features into products and internal tools
  • Solution architects assessing agent-based architectures
  • Technical teams looking to transition from prototypes to production environments

Number of participants


Price per participant

Testimonials (3)

Upcoming Courses

Related Categories