Introduction to agentic software engineering

This half-day online workshop introduces the principles and practical applications of agentic AI in software engineering. Through expert guidance and live demonstrations, you'll learn how AI agents work, how they differ from traditional AI coding assistants, and where they can add value in software development.

Led by Dr Panos Patros, the workshop combines practical insights with demonstrations covering requirements analysis, software architecture, development and debugging. You'll also explore how to work effectively with AI agents while maintaining appropriate verification, human oversight and professional accountability.

The workshop concludes with a panel discussion exploring how agentic AI is reshaping software engineering practice and what this could mean for the future of the profession.

Intended audience

This workshop is designed for software engineers, developers, technical leads and engineering managers interested in understanding how agentic AI can be applied in software development.

It’s suitable if you’re:

• beginning to explore agentic AI
• already using AI coding assistants and want to build your understanding of more advanced capabilities
• interested in how AI may influence software engineering processes and practices.

No prior experience with AI agents is required, although a basic understanding of software development practices will be helpful.

Learning outcomes

By the end of the workshop, you'll be able to:

• explain what AI agents are and how they differ from conventional AI chatbots
• understand how agent operating modes, including planning, execution and debugging, support software development tasks
• recognise the importance of context engineering, including project instructions, rules and tools, in improving agent performance
• identify opportunities to apply agentic AI across the software development lifecycle
• understand the importance of verification, security, human oversight and professional accountability when working with AI agents
• recognise how agentic AI may influence software engineering processes, tools and team responsibilities.

Course format

This four-hour online workshop combines introductory concepts, live demonstrations and opportunities for questions and discussion.

You'll see agentic AI applied to different stages of software development, including:

• Requirements analysis and elicitation: using AI agents to explore user needs, clarify assumptions, and develop structured software requirements.
• Architectural planning and construction: using AI agents to support software design, planning and implementation, including identifying appropriate testing and verification requirements.
• Autonomous debugging:  using AI agents to generate and run tests, investigate failures, identify potential causes and propose fixes.

Throughout the workshop, Panos will share practical techniques for managing context, selecting appropriate agent modes and reviewing AI-generated outputs, with a strong focus on testing, verification and software quality.

The workshop is demonstration-led rather than a guided hands-on session.

If you'd like to follow along, set up an AI coding agent before the workshop. Options include Cursor, Cline, GitHub Copilot and Claude Code.

Panel discussion

The final hour features a panel discussion with Fabian Gilson, Sean McGrath and Kenneth Johnson.

Together, they'll explore how agentic AI may change software engineering processes and practices, including implications for code review, verification, software quality and the evolving role of the engineer. You'll have the opportunity to ask questions and hear a range of industry and academic perspectives.

Presenter and panellists

The presenter and panellists contributed to the development of Engineering New Zealand's AI guidance for software engineering.

Dr Panos Patros | Workshop presenter

Panos is a Principal Engineer at Serato, where he leads the AI Development Transformation Group. He holds a PhD in Computer Science and has extensive experience in software engineering research, education and industry leadership. His previous roles include VP of Engineering at Raygun, Senior Lecturer and Deputy Head of Software Engineering at the University of Waikato, and Research Software Engineer at IBM.

Associate Professor Fabian Gilson | Panellist

Fabian is an Associate Professor in Software Engineering at the University of Canterbury. His research focuses on applying AI and machine learning to software engineering, with interests spanning agile development, software architecture, DevOps and testing. He brings experience across academia and industry, with a focus on improving software development practices through research and innovation.

Sean McGrath | Panellist

Sean is an AI Solutions Architect at Holmes Group, with a background spanning systems engineering, software development, cybersecurity and cloud infrastructure. He develops practical AI tools and agentic software workflows, alongside mentoring teams on using AI effectively to improve their work and processes. His focus is on keeping pace with rapid advances in AI and translating emerging capabilities into useful, reliable solutions.

Associate Professor Kenneth Johnson | Panellist

Dr. Kenneth Johnson is an Associate Professor at AUT and researcher in trustworthy AI and software engineering. His expertise spans formal verification, self-adaptive systems, and the engineering of reliable AI-based technologies across domains including disaster management, public health, and cloud computing.