Waitlist: Get Better Results from AI Agents in One Day (For Non-Technical Professionals)
Interested in the next workshop? Register here to hear when new dates are announced.
You've been using AI chat for a while now.
You've written the prompts, refined the answers, and got real work out of ChatGPT or Claude Chat. And yet:
You hit your usage limits far sooner than expected. You wait for the reset or pay for more, and the results don't get proportionally better.
You got a brilliant result once and can't get it again. The client briefing that landed perfectly last month comes out generic this month, and you can't tell what changed.
Your output is fine, and stuck at fine. AI gets you 60% of the way on a report or a deck, and no amount of nudging gets it to the 90% you'd actually send.
You spend more time correcting than time saved. The draft looks plausible, then you notice it ignored a key requirement or assumed the wrong audience, and you're on round four of fixes.
Your setup works, until something breaks. A saved prompt stops behaving, a project file gets overwritten, and you can't find the version that worked or remember how you fixed it last time.
You redo the same task from scratch every time. The weekly report, the research digest, the document review against a checklist: each run starts with a blank prompt instead of a process the AI already knows.
You've automated pieces of your work but nothing joins up. A prompt here, a template there, and none of it works with your actual files or hands off to the next step.
You use AI well for writing, and suspect that's the shallow end. You want it doing research, analysis and review across your files, not just polishing prose.
If you lost your laptop tomorrow, your AI setup would go with it. Months of custom instructions, saved prompts, project files and working documents, with no way to restore them.
If any of these sound familiar, you've outgrown the chat window.
Each has a specific, learnable fix, and that's the workshop.
Meet your trainer
Gang Rui (LinkedIn) built production agent systems before tooling reached the mainstream. He is one of Singapore's three founding Claude Ambassadors and has taught these methods to more than 50 professionals, from operations teams and executive assistants to PR staff and founders. He teaches in the language of your job, so you get the leverage of the best coding principles without having to learn to code.
Why this workshop is different
Claude Code and Codex were built for software engineers. The people getting the most out of them are the engineers who know how to manage context, break work into phases, version their workspace, and package what works into reusable skills. Those are the practices behind every reliable agent. This workshop takes those disciplines and teaches them in a commercial context to professionals without a coding background.
It is not a demonstration, a list of prompting tricks, or help guides you could read on the Claude or Codex pages. Everything in the room is Gang Rui's own operating model: the skills, workspace setup and working habits he built and refined on real client work, packaged so you can install and run them the same day.
You bring one task from your job that produces a recognisable result and that you expect to repeat: a recurring report, a research briefing, reviewing documents against criteria, turning raw material into a client-ready artifact. During the day you complete it with your chosen agent inside his workspace setup, guide it until the output is genuinely useful, and capture the method as a skill you can run again next week.
By the end of the day you will have:
Gang Rui's starter kit of skills and best practices, installed in your own workspace;
his operating model for running agent sessions, so more of your usage goes to useful work;
one high-quality reusable AI workflow from your own job, packaged as a skill;
a safer workspace with a version history you can recover from; and
an understanding of how reusable workflows can power automation of recurring work.
And it doesn't end after the workshop.
Every participant gets lifetime access to our private WhatsApp group: ask Gang Rui directly when you are stuck, and keep building alongside a community of professionals who came with real use cases. AI moves fast, and the group is where you stay ahead of it together.
Format
Participant Cap : 15 participants maximum.
Tools: Claude Code or Codex. Both are demonstrated; the principles are identical.
You need: a Claude Pro plan (for Claude Code) or a ChatGPT Plus plan (for Codex), and a laptop you can install software on.
Level: This is not an AI introduction course. You should already use an AI chat app comfortably and have attempted refining its answers with follow-up instructions. No coding experience is required.
What you will learn
The four modules are the four parts of his operating model. Each is a discipline engineers use with coding agents, stripped down to what a non-technical professional actually needs.
Manage your sessions so tokens go to useful work (what engineers call context management)
Get consistent quality on larger tasks (phased execution)
Keep your workspace safe and recoverable (version control)
Package what worked into a reusable skill (a proprietary 6-step framework)
What past participants say
Frequently Asked Questions
I've done AI workshops before. What's different?
Most show you features and send you home with notes. Here you leave with your own task running as a skill in your own workspace, built with the operating model Gang Rui uses on actual client work.
I have no coding or engineering background. Is this too technical for me?
No. Gang Rui built his operating model so that someone with no coding knowledge can set it up and run it. The class is kept small so he can make sure everyone's setup is working before they start the build in class.
How do I know if I am ready for this course?
If you use ChatGPT or Claude Chat comfortably and refine its answers with follow-up instructions, you're ready. This workshop suits anyone who wants more out of their AI setup and is willing to iterate to get there.
Is this a coding workshop?
No. You'll learn the principles engineers use to get reliable work out of agents, explained in plain terms, but you won't write or need to know any code.
Can't I learn this from the docs, or by asking Claude?
The docs tell you what should work in theory. Gang Rui tells you what works in practice. He has tested these practices on real client work and uses them every day. He also surfaces the problems the docs don't cover and gives you workable solutions, so you get his current, tested setup instead of figuring it out yourself.
Why not just keep using the chat app?
Chat answers one question at a time. An agent works with your files, follows a process, and runs a skill without you re-explaining it. That's why your good results become repeatable.
Claude Code or Codex?
Whichever you already pay for. Both are shown in the room and everything taught applies to either.
What if I fall behind in class?
That's what the 15-person cap is for. Most of the day is hands-on building with individual help, and you'll get direct guidance on your task.
Will the workshop be recorded? No. It's live and hands-on, and the value is in building your own task with help in the room, so plan to attend the full session.