We teach you to build it, not build it once for you.
We teach business owners, in person, the same way of working we built for ourselves from scratch. Afterwards, your own team keeps changing and growing the Hub. You don't end up with one more vendor whose schedule you have to wait on.
What we deliver is an Intelligence Hub
As a company grows, disorder happens on its own: experience walks out with people, everyone describes things their own way, and processes slowly drift. We believe digital transformation is, at its core, a continuous fight against this organizational entropy. The Intelligence Hub is the way of working Puchuang AI builds with a company, so its people and AI work together. We write down the company's experience, its standard answers and its processes as documents AI can read and follow, and everyone works from a single entry point. Once the Hub is in place, every task an employee does draws on the whole company's knowledge. Founders can use the outside view AI brings to look again at their organization and business, and make the final call based on their years in the industry. Build the brain first, then grow the hands: roles, tools and automation all grow around the Hub, instead of piling up into a mess.
While it's being built, you're there the whole time and walk through every step yourself. Once it's built, your own team maintains it. That's what we mean by teaching you to build it.
Read the essay: Build the Brain FirstHow the services connect
You can start anywhere. Writing one role down is configuration; writing the whole company down is the Hub; once it is written down, the part you hand to AI is an agent; what grows from there are the solutions.
Why we teach you instead of doing it for you
Doing it for you once only solves it once
Your business keeps changing. Whatever is configured today has to change every time the business does. If every change means going back to someone else, your system keeps growing in their hands. Whether it can change, and when, is no longer up to you.
You're not buying hours. You're buying certainty.
The hard part was never the doing. It's knowing which way to go. A method can be boiled down to one sentence precisely because getting there is hard. Most people get stuck not knowing how to break things into parts, not knowing what comes next, trying again and again without getting through. What we sell is a shorter path through that stretch.
Two paths, side by side
We offer two paths. On one, we configure it for you and hand it over. On this one, you configure it yourself while we guide you. If you need something working fast, choose the first. If you want to be able to change it yourself later, choose this one. Both are available, neither comes first, and neither replaces the other.
AI Configuration includes team training, but there it's one step on that path: we configure, then teach your people to use it. This path is different. From day one you do the work yourself, and we walk alongside you.
Choose a depth, then go through four stages
Not everyone needs to go equally deep. Before we start, we settle one question: do you want to learn to explain the work clearly, or to make changes with your own hands?
Two depths
Explain the work clearly
Learn to turn what needs doing into tasks AI can take on, organize company materials so AI finds the right thing, and check the work it hands back. No technical details, but you're in charge from start to finish.
Make changes yourself
On top of that, learn to bring things onto your own computer, change them, and publish the changes yourself. Suits founders who want to get hands-on, or teams with someone who can take it on.
Four stages
Diagnose
First we get clear on which parts of your business are worth doing, in what order, and which depth suits you. You get a written proposal.
How we know it's done
After reading it, you can say which part to start with and why. If you can't, this stage isn't done.
Lay the foundation
We set up the working environment and build the first business module that actually runs. You walk through it yourself from start to finish, rather than watching us do it.
How we know it's done
Without us there, you can walk through it again and it still works.
Build capability
Following the order set in the diagnosis, we build one module at a time. Along the way, the process becomes your own knowledge base and working rules, so whoever takes over next can simply follow them.
How we know it's done
Someone who wasn't involved at all can take over one module by following the rules, and keep it running.
Run on your own
You keep improving it yourself. We're available as advisers when you need us, and when you don't, there's no need to keep us around.
How we know it's done
Over a stretch of time with no involvement from us, you still build the next module.
The end of every stage is an exit. If you feel the direction is wrong, or you can't afford the time right now, you can stop there. The Hub configuration and knowledge base delivered so far are yours. We don't take them with us.
This path isn't right for everyone
Better to say it plainly up front than to have both sides regret it later.
Good fit
- The business is stable and can carry a stretch of investment before results show, without panicking over short-term outcomes
- Parts of the business can be turned into processes. Without processes, there's nothing worth building
- The founder wants to be hands-on, not send someone else to sit in
- You agree with "a bit slower, but it grows in your own hands"
Not a fit
- You want to buy a ready-made system. That's outsourcing, not this path, and we'll point you straight to AI Configuration, which is faster
- The founder won't take part. When staff are sent to learn and the founder stays out of it, the decision chain breaks in the middle and the learning doesn't stick
- The business is highly non-standard and barely runs online. There's too little to digitize for the investment to pay off
We go through these face to face during the diagnosis. If it's not a fit, we'll say so. A job that's a poor fit becomes a burden for both sides once it's done.
We use this method every day
Our customer management system, business dashboard, customer support tools, and the website you're reading now were all built with this method, and we change them every day.
So everything we teach is something we learned the hard way, including the steps we got wrong and why we changed course. It's not methodology copied out of a book.
Four questions first
No need to wait for a diagnostic. See for yourself where your company stands today. For each question, pick the line closest to you.
Is your knowledge written down?
Products, customers, house rules: where do they live now?
Do decisions leave a trail?
For what you settled last month, can you still find why?
Can changes be traced?
How many times has a process changed, who changed it, and why?
Is AI part of daily work?
How does your team use AI today?
What would you most like to fix first?
Pick as many as you like. If none fits, tap the last one and say it in your own words. Optional.
We'll ask these four questions again, face to face, during the diagnosis.
Getting started
A first conversation
Tell us where your business is today and which parts give you the most headaches. This step is about judging whether there's something worth doing.
A diagnosis
We look at how your processes actually run, work out what can be done and in what order, and give you a written proposal.
Decide with the proposal in hand
How many parts, which depth, how much to invest: it's all in the proposal. Read it, then decide whether to start.
We don't list prices online. Every business differs in complexity, in how many parts to build, and in which depth fits. All of that comes in the proposal.
FAQ
The brain is built. Which hands next?