If I Just Had the Confidence, I'd Do My Own AI in the Business.
You're not alone. I hear that a lot. We've all tried ChatGPT, Claude and Gemini. But the inbox, the quotes and the paperwork still land on your desk.
By John Hynds · November 16, 2026
You're not alone. I hear that a lot. We've all tried ChatGPT, Claude and Gemini. Maybe you've even set up a local bot or two (or 20). But the inbox, the quotes and the paperwork still land on your desk.
The tools are already on your desk
The tools are in your browser, in your office software, and on your phone. Some of you have already stood up an assistant or an agent and given it a job. The new AI agents, like Grok Bot, are powerful enough to act on your behalf. That power is potentially very dangerous without the right guardrails.
So the real question is simple: can I trust this with real work?
Because the real work has consequences. An email goes out under your name. A quote goes to a customer. A schedule gets changed. A file with client information gets opened. If you've only used AI for a first draft here and there, handing it your inbox feels like a leap.
So the tools sit there. You use them when you remember. And the busywork that eats your week stays right where it is.
I get it. I'd feel the same way if someone told me to point an unproven system at my payroll or my customer list and hope for the best.
A lot of owners would put AI on that work tomorrow, if they could trust it. They want to know it will stay inside the lines, and that they can run it when nobody is sitting next to them.
What changes when we build it together
Confidence comes from building. We sit down together, build a real system on your real work, and you leave knowing how to run it. Another demo or another list of prompts won't get you there.
That's the consulting. We sit down and work on the actual task that's eating your week. You build it in the tools you already use. You see why each step works. You see the guardrails go on before anything can touch a customer. By the time we're done, you can run it, change it and add the next piece yourself.
The training happens because of the consulting. You do the work, with me there, until you own it.
Here's how it usually goes.
We start with your busywork. We find the task you keep meaning to fix. Inbox triage. Quotes. Scheduling. Paperwork. Something concrete, so you can see it working on your own operation.
We build it together. You do the building. I sit beside you and we set it up the right way, with guardrails, so you understand each step well enough to fix it later.
You own it. It runs in your accounts, not mine. It's yours. You can keep going with me or run it on your own. That's up to you.
When you've built it yourself, the confidence sticks. You know why it works. You know what it can and cannot do. And you're not stuck waiting on someone else when you want to change it.
You stay in command
You're putting business data in front of AI, so you need to know what it can touch.
The systems we build run in your own accounts. They reach only what you authorize. They don't delete anything. A person approves anything that commits you: an email, a quote, a change that a customer will see. Everything is recorded.
You set the guardrails. AI does the work. You stay in command.
That's how the confidence gets built: you watch your own work run inside rules you can see and change.
Why it matters who sits beside you
A lot of AI help comes from someone who learned the tools last year. The tools matter, and so does judgment: what to build, what to skip, and how to keep it secure.
I've spent more than 40 years building and running things.
In 2003 I opened LoCarb Outlet with no retail experience. We did over $500,000 in sales in year one, and I made payroll. I've signed the front of a paycheck, and I know what it feels like when the week depends on the work getting done.
I also build software people pay for. I built and run CrewForge, a field-service app that paying customers use every day.
I worked Boeing's 747 and 737 production lines, and the manufacturing line at Genetic Systems, building the first in-field AIDS test systems in FDA-regulated manufacturing. I built Field Service there from the ground up.
I built Xylan's global sales engineering team through its IPO and its $2 billion acquisition by Alcatel, spent 13 years at Polycom, and co-authored patented optical network architecture as CTO of ClearStream. I led Sales Engineering at Device Authority, a zero trust cybersecurity company.
I'm a US Marine Veteran, an MIT xPRO AI Facilitator, and I hold an MBA from the Acton School of Business.
When we sit down together, you get someone who has run operations, led engineering teams, owned a business and shipped software that has to work on Monday morning.
That's the point. When you put AI on real work, you want someone beside you who has carried real responsibility before.
Let's talk. Coffee or a call.
If you want the confidence to do your own AI in the business, start with a free Risk Management Assessment. Coffee or a call.
Tell me where your week goes. I'll show you what you could build, with real screenshots, and we'll see if it fits.
Humans for purpose. AI for tasks.
Get new posts by email
Get the latest Hynds AI blog posts emailed straight to you. No spam, unsubscribe anytime.
Let's talk. Coffee or a call.
No pressure. I'll share what I've built, with real screenshots, and we'll see if it fits how your business runs.
