AI Engineer Workshop：Claude Agent Workshop

A hands-on day of building AI agents with Python—explore tool use, workflows and evaluations through practical coding.
About this Event
Claude Agent Workshop

A hands-on day of building AI agents with Python, exploring tool use, workflows and evaluations.

Build AI Agents & LLM-Powered Applications with Python.

With today’s models and development tools, building an AI application that answers questions is increasingly accessible. Giving it the ability to call tools, retrieve information and complete tasks requires an understanding of how agents work.

How does an agent decide what to do next? What happens after a tool returns a result? When should you use a predefined workflow, and when should the agent choose its next action? How do you know whether it completed the task correctly?

Understanding these questions—and exploring them through code—provides a foundation for developing, debugging and improving AI agents.

In this Claude Agent Workshop, Roy, AI Engineer at V2 AI, will guide you through practical Python coding and focused exercises.

Using the Anthropic Python SDK, you’ll start with a basic agent loop, then explore tool use, workflow control, external knowledge retrieval and evaluations.

🛠️ What will you learn in one day?

The workshop covers four core areas:

• Agent Development | Build with Python
  Use the Anthropic Python SDK to make model calls and implement a basic agent loop. Follow how a user request moves through model responses, tool execution and returned results.
• Tool Use & Workflows | Connect tools and control execution
  Define tools, validate arguments, handle execution results and return them to the model. Compare predefined workflows with agent-directed execution, where the model selects tools and determines its next action.
• MCP & RAG | Connect external tools and knowledge
  Work through a prepared MCP integration and a small document retrieval exercise. Explore how agents access external capabilities and use relevant information to support their answers.
• Evaluations & Debugging | Test and improve your agent
  Assess task completion, tool selection and answer quality. Run a small evaluation set, inspect failures and adjust instructions or tool behaviour based on the results.

The workshop centres on practical work. You’ll begin by running and inspecting an example, then write code, connect tools, test results and make improvements—bringing the concepts together in a runnable agent project.

🎯 What will you take away?

The workshop connects technical understanding with practical experience in three ways:

• Experience building an agent from scratch
  Develop a runnable Python agent project and understand how model calls, tool execution and the agent loop work together.
• A clearer understanding of the development process
  Connect tool use, workflows, MCP, RAG and evaluations. Understand the problem each component addresses and how the pieces support task execution.
• Code and exercises to continue learning
  Take away workflow examples, guided MCP and RAG exercises, and a small evaluation set with observations to inform further improvements.

You’ll leave with hands-on experience across the key stages of agent development and a clearer sense of what to practise next.

Who is this workshop for?

This workshop is suitable for:

• Learners with Python foundations who want a practical introduction to agent development.
• Software developers exploring LLM applications and AI agents.
• Developers who have experimented with agents and want to better understand tool use and execution flows.
• Aspiring AI engineers seeking practical experience with MCP, RAG and evaluations.
• Technical learners looking to connect individual AI concepts through a working project.

You should be comfortable reading and writing Python and ready to modify, run and debug code during the workshop.

Please bring a laptop with Python, uv and a code editor such as VS Code installed. The required API provider, supported Python version and responsibility for API usage charges will be confirmed before registration.

Meet your instructor: Roy

AI Engineer @ V2 AI

Roy has worked on Agentic AI and software development projects at companies including Deloitte, bringing experience applying AI engineering to enterprise needs.

His project experience spans global supply chain risk management, invoice processing automation and software development productivity, with a focus on connecting AI capabilities to practical business requirements.

Roy holds the following Claude certifications:

• Claude Certified Architect – Professional
• Claude Certified Architect – Foundations
• Claude Certified Developer – Foundations

Event details

Date: Sunday, 15 November 2026
Time: 10:00 AM–4:00 PM AEDT, proposed
Location: Melbourne, Australia; exact venue to be confirmed
Format: In-person, hands-on workshop
Language: English
Early bird: A$328 per person
Standard: A$389 per person

Early bird registration closes on 8 November 2026 at 11:59 PM AEDT, Melbourne time. Each ticket covers this workshop only.

Register for the workshop

Turn your understanding of AI agents into working Python code.

From building an agent loop to connecting tools, retrieving knowledge and evaluating results, spend a day putting the foundations of agent development into practice with Roy.