Build real software with AI.
For software developers and anyone who wants to build real software with AI.
One engineer can now build what a department used to. Real software, autonomous agent systems, AI automations. The practice, taught by an engineer who actually ships.
1:1 access included · Monthly workshops · Cancel anytime
Builders inside the community work at
You're already using AI. But you want to go further.
You've used Claude Code or Codex. You've shipped a few things. You can feel something has shifted. One engineer with the right setup can now do what a team used to take a quarter to deliver.
But you also feel the gap. Between throwing prompts at an agent and actually engineering with it. Between a demo that works once and software that holds up in production. Between knowing a few tricks and having a real practice you can apply to anything.
This is the gap AI Engineer closes.
What you'll be able to do.
Build real software, end to end.
Using AI coding agents as your workflow. From idea to a working thing other people can actually use.
Build production-ready AI agents.
Agents that hold up in front of real users. Built with modern agent SDKs, deployed properly, monitored, evaluated.
Build AI automations that run themselves.
Software that does operational work continuously, not just on request. Real engineering, not no-code duct tape.
Apply AI inside your real work.
Automate the repetitive parts of your job. Build agents that handle the boring middle of your workflow. The kind of system you'd want even if no one else ever saw it.
Stop chasing tool tutorials. Master the workflow once. Apply it to anything new that comes along.
The methodology is built around principles that survive tool churn. Claude Code today, something else tomorrow. The way you engineer with AI is what compounds.
A community for serious technical builders.
Working engineers, technical leaders, self-taught coders. Anyone writing real code who wants to build serious AI software.
Engineers and developers
You want to upgrade how you ship. Workflow, agents, production systems, the real work.
CTOs and technical leaders
You need to operationalise AI in your team and you want to do it on real engineering foundations.
Researchers, scientists, and operators who code
You're not building products. You're building the tools and workflows your real work needs. Literature digesters, lab automations, internal systems your team will actually use.
Self-taught builders
You learned Python to do your job better. You can already ship things. You want to close the gap between what you're doing now and what's actually possible with agents.
The filter is technical commitment. Not credentials. Not job title. Not whether you do this for money. You read and write code, especially Python, and you want to build AI systems that hold up in production.
If you don't write any code yet, start with the YouTube channel. You'll find your way back here when you're ready.
This isn't another Claude Code course.
- 01Tutorials that work in the demo and fall apart in production.
- 02Hype videos that teach features instead of judgment.
- 03Frameworks that hide what’s actually happening under the hood.
They teach features. They don't teach principles. Knowing when an agent's output is good and when it's plausible-looking junk. Knowing when to let agents drive and when to take the wheel. Knowing how to design AI systems that hold up in front of real users and real data.
That's the craft. Anyone can learn it. It just takes someone who has built real software for 20 years showing you how.
Tools change. Principles don't.
Built by Owain Lewis
20-year software engineer. Self-taught.
I'm an outsider to traditional “guru” culture. I came up self-taught. So have most of the people inside this community. Anyone can learn this. You just need someone who has actually done the work showing you how.
Founder of Gradient Work, an AI consulting agency shipping production AI systems for clients.
I've worked with hundreds of engineers from Google, Microsoft, Amazon, Netflix, and dozens of startups. I've been an engineering manager and director. I've owned and operated large-scale cloud services. I've built software you've probably used, and I've watched the entire industry get reshaped twice. I'm watching it happen again with AI.
I built AI Engineer because the existing AI education space is mostly hype and surface-level tutorials. I wanted to teach the real craft. The workflow, the practice, the patterns that hold up. In a way that anyone serious about learning it can actually pick up.
- 20+ years shipping production software
- Worked with hundreds of engineers from Google, Microsoft, Amazon, Netflix, and dozens of startups
- Engineering manager and director. Owned and operated large-scale cloud services
- Founder of Gradient Work (AI consulting agency)
- Teaching what I'm learning, every week, on YouTube
Two flagship courses. One membership.
Everything you need to engineer with AI. The workflow, the production systems, the community, and direct access to me.
Agentic Engineer
The disciplined way to build software with AI coding agents, drawn from 20 years of writing software. Works with whatever you're using: Claude Code, Codex, Cursor, Copilot, Antigravity. How to plan before you build. How to spot when an agent is bullshitting you. How to ship something that holds up. How to build your own agent.
Capstone: Build something you'd actually use. A web app, an agent that runs continuously, a workflow tool for your team. I demonstrate by building a content workspace. A researcher in the community is building a literature-digestion agent. A CTO is building an internal dashboard. The shape is yours.
AI Architect
Build AI agents and automations that hold up. In front of real users, real data, or your own work. The technical depth is there (retrieval, evaluation, cloud deployment) but you learn it by building something that matters to you, not by reading theory.
Capstone: Build and deploy a document Q&A agent. A system that answers questions about your own documents. Python backend, Next.js frontend, proper evaluation and monitoring, deployed to the cloud.
Monthly Workshops
Once a month I run a live workshop. Building something real, walking through production engineering decisions, demonstrating the patterns in real time. Members join live, ask questions, and shape what we cover. Recordings stay in the community.
The Skills Vault
Templates, prompts, the agent configuration files I actually use (CLAUDE.md, AGENTS.md, custom skills, etc.), review checklists, planning templates, and deployment configs. Copy-paste starters for your own work. Updated as the field evolves.
Community + Direct Access
A supportive community of working engineers and serious builders shipping real AI software. Ask questions. Share what you're building. Get unstuck. Direct async access to me. Not 'office hours.' Not gated.
Here's what happens when you join.
- Start Agentic Engineer. Set up Claude Code or Codex the right way.
- Apply your first planned workflow to something you're already working on.
- Introduce yourself in the community.
- Begin AI Architect. Learn the foundations of building with agent SDKs (the modern way to wire agents into your own code).
- Build your first real agent.
- Start the RAG modules (giving an agent access to your own documents).
- Start building toward the document Q&A capstone.
- Attend your first Monthly Workshop.
- Ship something real you'd be confident putting in front of a user.
After 30 days, you have working knowledge of the practice, real production patterns in your toolkit, at least one real artifact you've shipped, and you're part of a community of engineers doing the same work.
One price. Everything included.
Monthly is the default. Annual saves about 33% off the monthly rate and locks in your price for the year.
Pay as you go. Cancel anytime.
- Agentic Engineer. Flagship course
- AI Architect. Flagship course
- The Skills Vault. Templates, prompts, agent skills, patterns
- Monthly Workshops with me
- Async community access
- Direct 1:1 access to me
14-day full refund
Save ~33% vs monthly. Equivalent to $33/mo.
- Everything in monthly
- Locked-in rate for the year
- Save ~33% vs paying monthly
14-day full refund
Frequently asked.
I'm not a software engineer. Will this be too advanced?+−
I'm not a developer at all. Is this for me?+−
I'm a senior engineer or CTO. Is there real depth here?+−
Should I go monthly or annual?+−
Will the courses stay current as AI evolves?+−
How much time does this take?+−
Can I get a refund?+−
The path to mastering AI coding.
Build real apps. Make money. Grow your career. Become irreplaceable. With a community of working builders behind you, and someone who has done the work showing you how.
Anyone can learn this.
Monthly or annual. 14-day refund. Cancel anytime.