Twenty-eight questions. Seven levels. A take-it-seriously self-assessment
that tells you exactly where you are and what to work on next.
Free Access
The AI Skills Ladder
A 28-question self-assessment across 7 levels. Get an honest map of where you are on the AI skills ladder and a clear next step.
Most professionals overestimate their level by exactly one. The gap
between levels is not knowledge. It is proof of work.
Answer each question honestly, not aspirationally. Each level has four
questions. The goal is not to feel good. The goal is an honest map.
How often do you use AI tools in your daily work?
When an AI gives you a wrong or low-quality answer, what do you do?
Have you joined any AI communities, forums, or Discord servers to follow the field?
Have you completed any structured AI challenge (30-day challenge, cohort, or course)?
How would you describe your prompting approach?
Do you have a personal prompt library you actively maintain and reuse?
How do you evaluate whether an AI response is actually good?
Have you tested the same prompt across multiple models (Claude, Gemini, GPT-4.1) and documented the differences?
How many active automations do you currently have running that use AI?
Have you published any automation workflows or templates publicly (GitHub, n8n community)?
Have you participated in a hackathon focused on AI or automation?
Have you contributed to any open source automation project (n8n, Zapier, Make)?
What best describes your GitHub presence?
Have you built and published an MCP (Model Context Protocol) that others can install and use?
Have you built a RAG pipeline, agent, or AI-powered tool that at least 10 people actively use?
Have you contributed to major open source AI repositories (LangChain, LlamaIndex, AutoGen, Hugging Face)?
Have you led an end-to-end AI initiative with documented, measurable business outcomes?
Do you have a portfolio of AI work that includes outcome data, not just project descriptions?
Have you designed and delivered a training programme on AI for a team or community?
How many people have you actively mentored in AI skills (not just advised once)?
Have you designed or contributed to an enterprise multi-agent AI architecture in production?
Have you built a platform or framework that other developers or teams build on top of?
Do you run or actively contribute to a developer community, Discord, or open source project with regular contributors?
Have you presented AI architecture work at a technical conference or published a formal architecture document?
Have you filed for or been granted any AI-related patents?
Have your AI frameworks, models, or methods been cited by others in publications, reports, or public work?
Have you created a category, defined a new methodology, or launched an open source project with a community that grew independently?
Do you advise, keynote, or contribute to standards bodies, government initiatives, or industry consortia on AI?
Your Level
Your climb path
The Proof of Work principle
The gap between levels is not knowledge. It is proof of work. You do
not climb by learning more. You climb by shipping more. Prompts,
automations, tools, MCPs, case studies, platforms, frameworks, and
the people you train along the way. If you cannot point to the
evidence, you are not at that level yet, no matter how much you know.
Enter your email and I will send your personalised AI Skills Ladder report plus weekly practitioner notes. No spam.
The 7 levels in full
Level 0
The Dabbler
You use AI reactively. Paste a question in, copy an answer out. No system, no consistency, no way to judge if output is good or just plausible.
Proof of work: You can describe the AI tools you use.
Tools now: Claude Haiku, Gemini Flash, ChatGPT, NotebookLM
Level 1
The Practitioner
You prompt with structure and intention. You know output quality is a direct function of input quality. You direct AI behaviour rather than ask it questions.
Proof of work: A documented prompt library. Consistent output quality.
Tools now: Claude Sonnet, Gemini Pro, GPT-4.1, Perplexity, personal prompt libraries
Level 2
The Automator
You stopped doing manually what AI can handle. You build workflows that run without your constant attention. You have reclaimed hours per week.
Proof of work: Live automations with measurable time savings. Workflows others can see and test.
You do not just use AI tools. You build them. You design systems where AI does the reasoning and you architect the surrounding logic. Others depend on what you create.
Proof of work: GitHub repos with stars and forks. Published MCPs. Hackathon submissions. Apps others use.
Tools now: Claude API with tool use, LangGraph, LlamaIndex, Cursor, Replit, Pinecone, Supabase
Level 4
The Strategist
You translate AI capability into business value at organisational scale. You decide where AI should and should not be applied. You design for outcomes, not features.
Proof of work: Portfolio of AI initiatives with outcome data. Published frameworks. Training programs delivered.
Tools now: Gartner AI Maturity Model, NIST AI RMF, McKinsey State of AI, EU AI Act literacy
Level 5
The Architect
You design enterprise-grade AI systems that operate at scale, under governance, across complex organisational and technical constraints. You think in systems, not features.
Proof of work: Platform architecture that others build on. Technical papers. Active mentorship program.
Tools now: LangSmith, W&B, Azure AI Foundry, AWS Bedrock, OpenTelemetry for AI
Level 6
The Innovator
You create the intellectual frameworks others use to think, build, and govern AI. You do not follow the map. You draw it.
Proof of work: Patents. Citations. Communities built. Categories created. Keynotes that change how an industry thinks.
Tools now: Your own published frameworks, platforms, and research