Your design team's AI practice
doesn't have to be chaos
In five weeks, your whole team moves together. Same standards. Same quality. Same growth curve. Then the system keeps improving itself.
Programme Details
- Next cohort
- July 2026
- Time commitment
- 4 hours/week
- Duration
- 5 weeks
- Format
- Hybrid: e-learning + live workshops
- Structure
- Learn + Build
- Cohort size
- Limited (8–12)
TBA
Get updated on cohort dates before anyone else.
This is for people already fluent with Claude
You use Claude daily. You know how to iterate. You've hit the ceiling of individual productivity. Ready to scale it.
You don't need a technical background, but you need to be comfortable directing a model and iterating with it. If you're still finding your feet with Claude, our AI Thinking Workshop Series is a better fit.
Scattered LLM use isn't a people problem. It's a systems problem.
Right now, every person on your design team has a different relationship with their LLM. Different prompts, different quality bars, different results. Your best person is getting remarkable outputs. Someone else is reinventing the wheel every single day and nobody can explain the gap, let alone close it.
The problem isn't the tools. Claude is extraordinary. The problem is that nobody has designed how the team uses it together. There's no shared grounding, no common understanding of your standards, your clients, your quality bar. Every conversation starts from zero.
And even when teams start building systems, they stop there. The system gets stale. Nobody audits it. The gap opens back up six months later.
This is the new Design Ops skill: learning to orchestrate an AI-augmented team and building the system that keeps improving itself.
A complete, working Agentic Design OS.
Not a strategy document.
Every session produces a concrete artefact. By week five, the pieces connect into a governed, self-improving system your team deploys and maintains.
Map where you actually are
Audit your team's current Claude use. Identify the gaps, inconsistencies, and hidden expertise worth codifying.
Give everyone the same rulebook
Build the shared context, mission, quality standards, and client knowledge that every Skill and Agent runs from.
Turn your team's best work into templates
Design purpose-built Skills that embody your team's practices. Build multi-step Workflows that connect them into repeatable processes.
Make your OS talk to what you use daily
Build Agents that use your Skills to make decisions and take action. Connect to the tools your team already uses.
Build the system that improves itself
Design the audit process, compounding learnings protocol, and improvement cycle that keeps your OS sharp over time.
Five weeks. Five live sessions.
One system that keeps improving.
Each week pairs a learning session with a building session. You arrive with a framework. You leave with something that works.
Foundations and what an Agentic Design OS actually is
The architecture of a team intelligence layer. Why scattered LLM use produces inconsistent results. The anatomy of a system that doesn't.
Audit your team's current AI use. Map your team's current state, the gaps, inconsistencies, and hidden expertise that should be codified.
Grounding and giving your OS the context it needs to be useful
Why grounding documents are non-negotiable. Mission, vision, ICPs, product knowledge, quality standards, competitor landscape.
Build your team's grounding document set. Draft and test the shared context that underpins every Skill and Agent you'll build.
Skills and Workflows turning your SOPs into AI-powered systems
How to design elite Claude Skills and multi-step Workflows. What makes a Skill genuinely useful versus generic.
Build your first Skills and Workflow. Map your highest-value, most-repeated design ops tasks.
Agents and MCPs connecting your OS to the tools your team uses
How Agents and MCP connections work and when to use them. How Agents use your Skills to make decisions and take action.
Build your first Agent and connect your first MCP. Design an Agent that uses your Week 3 Skills to handle a real repeatable task.
Evolution connecting it all and building the system that improves it
Governance, auditing and compounding intelligence. How to connect your grounding, Skills, Workflows, and Agents into a coherent governed system.
Complete your OS and build your evolution plan. Assemble all five weeks into a complete, governed, connected system.
For practitioners already deep in Claude
who are ready to build the system around it.
Design Ops Specialists
You think in systems. This is the next system, the AI layer that turns your team's individual Claude fluency into a governed, compounding team capability.
Heads of Design
You're accountable for quality across the team. This is how you close the gap between your best person's output and everyone else's, permanently, not occasionally.
Senior Designers
You're already getting strong results with Claude. This gives you the infrastructure, Skills, Agents, grounding, that makes your expertise available to the whole team.
Most AI courses teach individuals.
This one transforms teams.
| Other AI courses | This course |
|---|---|
| One person gets strong results | Your whole team matches your best person |
| Focus on speed and automation | Focus on quality that compounds over time |
| Generic frameworks to adapt later | Built around your actual team, standards, clients |
| Knowledge you apply eventually | Working system you deploy on Monday |
| A system you build once, then ignore | A system that gets smarter every week |
| When people leave, knowledge walks out the door | Knowledge stays. People are replaceable. Systems aren't. |
What designers say
This recasts AI beyond being a vending machine for cheap, easy answers to being a genuine partner. It takes a far broader perspective, asking you to consider AI from ethical, organisational, governance, and audit angles.
An abundance of applicable information that can be applied immediately, with a large impact. This course was a game changer, not just for how I work, but how I approach AI interactions and design.
This course was brilliant, hands-on tools, tips and tricks to use AI effectively while demonstrating how to use it responsibly. The interactive sessions help as you learn from people all over the world grappling with similar challenges.
The questions we always get.
Do I need to be a developer or know how to code?
No deep technical background is required, but you do need to be comfortable using Claude to help you build, write and debug. You'll be directing Claude to code components of your OS, reviewing and iterating on outputs, and connecting tools. This is well within reach for anyone with solid LLM fluency and a systems-thinking mindset.
What does "intermediate LLM fluency" actually mean?
It means you use Claude daily, you're comfortable with multi-turn conversations, you know how to iterate to get quality output rather than relying on single prompts, and you've worked with Claude Projects or equivalent features. If you're still in the early stages of getting comfortable with Claude, our AI Thinking Workshop Series is a better starting point.
Is this specifically for Claude, or does it apply to other LLMs?
The architecture, grounding, Skills, Agents, governance, evolution cycles, applies to any LLM. We use Claude because its Skills, Projects and MCP ecosystem is the most powerful and accessible for this kind of team infrastructure. But the thinking transfers directly.
My team is small. Is this still worth it?
Especially worth it. Small teams benefit most from a shared intelligence layer because there's no redundancy. When knowledge walks out the door, it's gone. This course helps you systemise what your best people know before that happens.
Every month you wait is a month your team isn't compounding knowledge.
The next cohort starts July 2026. Spots fill fast. Build your system while your team's still learning Claude — before knowledge walks out the door.
Get updated on cohort dates before anyone else.
Limited places per cohort.