Small, hands-on workshops in Bangkok
Bring one business problem. Leave with a working first version.
Practical AI workshops for owners and operators responsible for real outcomes. Map the work, choose an approach, build something bounded, and inspect the risks with direct support.
Move to inspect the systemIllustrated concept scene
How the workshop works
One real problem. Five practical steps.
The tools come after the problem is understood. Each session moves from a real business task to a limited version you can inspect and improve.
01 / Bring a problem
Start with work that matters to your business.
Choose a task that takes too long, produces errors, or depends too much on one person. Define who does the work, what information they use, and what a good result looks like.
02 / Map the work
Understand the process before choosing AI.
Identify the people, data, decisions, handoffs, costs, and risks. A clear map shows where AI can help, where fixed rules work better, and where a person must approve the result.
- Goal
- People
- Data
- Decisions
- Risk
- Cost
03 / Choose the approach
Choose models and tools for clear reasons.
Compare models, agents, workflows, skills, MCP servers, and existing components. Each choice is based on quality, cost, privacy, reliability, and maintenance.
04 / Build and test
Build a small version and test it.
Use your own laptop and a limited example from your business. Check the output, data access, permissions, failure cases, ownership, and the decisions that still need human approval.
05 / Keep control
Leave knowing what you built and what to do next.
When the problem is suitable, you leave with a tested first version, written requirements, known limits, and the next steps needed to make it safe and useful.
The Workroom.AI promise
Trust should come with clear rules.
If I ask for your trust, time, or money, these are the rules I will follow.
No fake sales activity.
I will not cycle fake names, purchases, attendee counts, or “recent activity” through JavaScript or any other system to make demand look real when it is not.
Free means free. Paid means one price.
If an event is free, I will label it free. If it is paid, everyone pays the same published price. There are no VIP, premium, or pay-more-to-get-the-real-experience tiers.
There is no inner circle.
I may create groups, Discord channels, or other ways to stay connected. They will be open to everyone, not reserved for people who pay more.
You can still ask me for help after the event.
If you have attended before and get stuck, ask me. I will help when I reasonably can. You will not need to buy access to a special group first.
Consulting will have clear terms.
If I offer consulting, I will publish the scope, price, assumptions, and responsibilities before work begins. You will know exactly what you are paying for and what you can hold me accountable for.
Why I run it
Bangkok is part of my family. Honesty is the standard I want to set.
Why Bangkok
Bangkok is where I met my wife. Our daughter, who will be half Thai, is due in about two months. This is not simply a city I enjoy living in. It is part of my family's life and future.
Why The Workroom
I see a gap in AI education created by a lack of honesty: claims without evidence, limits left out, and costs or conditions disclosed too late. I want The Workroom to be a useful resource that sets the standard for being clear about what works, what fails, what I know, and what I still need to verify.
That standard should be visible in the terms I publish, the way I teach, and the corrections I make when I get something wrong.
- Evidence before claims
- Terms published clearly
- Mistakes corrected openly
Next small session
Tell me you want to take part.
I will send the date, location, price, requirements, and booking terms together when the next Bangkok session is ready.
No payment now. No countdown. No sales sequence.
Words you may hear in the workshopPlain-language definitions
- Model
- The engine that analyzes information or generates an output.
- Agent
- A model that completes several steps using tools and stored information.
- Harness
- The instructions, tools, limits, approvals, and recovery around an agent.
- Skill
- Reusable instructions for a specific kind of work.
- MCP
- A standard way for compatible AI clients to access tools and information.
- System
- The software, data, people, rules, and infrastructure working together.