Claude Code: An agentic software experience

Description

Agentic engineering seems to finally deliver on the promise of notable productivity gains from AI. But what does it actually feel like to work in a team where AI agents are members? How do you steer and direct them? What patterns and practices emerge when you have multiple agents working in parallel on the same codebase?

In this hands-on and immersive workshop, AI agents work alongside you as full team members. Building, testing, and documenting in parallel while you steer. You are not invoking AI as a tool. Instead, AI is your teammate: agents that take initiative, pick up tasks, and deliver results. Your role is to direct, review, and iterate.

You walk the whole chain, not just the coding part. An agent interviews you to map the user stories and turn them into specifications, you build those specifications feature by feature, a testing agent drives the application through a browser to check behavior and write end-to-end tests, and a dedicated review agent reviews the result and posts its comments back on the pull request. The agents connect to your issue tracker, wiki, and git repository, so the work lands where your team already works.

Every hour, a short knowledge pulse introduces a concept grounded in what you have just experienced: custom instructions, MCP servers, agentic patterns. Theory follows practice, not the other way around. Coaches are available throughout the day. You leave with working software and a clear sense of what agentic development feels like in practice.

Intended for anyone who builds software or wants to: developers, testers, product owners, analysts, and enthusiasts. Solo participants and teams alike. Organizations can enroll their own team and work in a context that is relevant to them.

Prior Knowledge

• Familiarity with software engineering
• Understanding of basic concepts within software development (features, tests, refactoring, documentation)