Agentic
Operator
A 4-week guided practice that takes non-technical operators from zero to shipping with Claude Code.
Through live classes and async material, participants set up Claude Code, build a real project from scratch, and develop the agentic mindset that separates AI consumers from AI operators.
- Kickoff: AI recap (probabilistic thinking, key concepts: model, harness, context). Course orientation and first setup.
- Lesson 1: Claude Code setup (IDE + GitHub). What you can build. AI Native Builder Mindset. Claude Code Essentials: prompting, CLAUDE.md, models and modes. First small project.
- Lesson 2: Working with unstructured data and content for broader projects. Claude Code Customizations: Skills, Plugins, and MCPs.
- Lesson 3: Building an actual product. How to start and structure a project agentically.
- Lesson 4: Tracking your work with an agentic project map. Agent-human parity and testing: knowing when to trust the output.
- Sets up Claude Code and starts a new project from scratch without hand-holding.
- Writes a CLAUDE.md file that guides the AI reliably across a project.
- Has shipped at least one real tool or product built during the programme.
- Applies the agentic-map mindset to track and evolve their own AI-powered work.
- Knows when to trust AI output and how to verify it before relying on it.
5 to 6 operators who can build first versions of internal tools without waiting for a developer. A team that configures digital workplaces for others can now configure AI for itself.