Hi, I'm Nick.
I help teams put AI to work properly.
Most teams start with a tool. I start with how your work actually moves and turn it into workflows on Claude, with your context written down once and agents for the repeating parts.
Illustrative example.
From the first question to a working system
Four steps. You can start at any of them, and each one ends with something your team keeps.
-
Strategy
We map your company's workflows and find where AI pays off, and where it should stay out.
You getA written plan in plain language: what comes first and what it costs.
-
Architecture
I design how the pieces fit: which tools, what context the AI sees, where a person checks the result and who looks after it.
You getA design your team, or any developer, can follow and maintain.
-
Training
Half or full-day workshops on Claude for the team and 12 weeks of 1:1 mentorship, all on your own documents.
You getA team that knows what AI is good for and what never to paste into it.
-
Build
I build what the plan calls for: agents and skills on Claude, linked to the systems you already use and tested on real work.
You getA working system, documented, and yours to keep.
Sometimes the plan says AI isn't the answer for a task. Then I say so, and that's part of the plan.
Most teams start with a workshop or a strategy call. The free call settles which one fits you.
Book a free callThe difference is in the request
A vague request gets a generic answer, and clear instructions with the right context get something you can send. Every training starts here.
Write an email to a client about the delay.
Dear [Client Name], I hope this email finds you well. I wanted to inform you that there has been a delay regarding [project/order]. We apologize for any inconvenience…
Blanks to fill in and a tone nobody uses.
Email to Maria at the hotel we supply. Her 40 chairs ship Thursday, not Monday, because the fabric came late. Offer free delivery. Short and friendly.
Hi Maria, quick update on your 40 chairs: our fabric supplier was late, so they'll ship on Thursday instead of Monday. Delivery is on us this time.
Ready to send after a quick read. Illustrative example.
Treat what AI tells you as a confident first draft. Check it before it goes anywhere.
AI always has an answer. Whether it's the right one is still your call.
Give a great model the wrong context and you get an average answer.
Start with one task. Once it works, add the next.
Spotting a wrong answer matters more than knowing where to get one.
AI writes what sounds likely. Checking what's true is your job.
A year of chats leaves you with chat history. A few saved skills leave you with a system.
Weak prompts hand control to the AI. Clear ones keep it with you.
What your team learns
Training follows three levels, each built on the one before. A workshop covers the first; mentorship covers all three.
Ask
Brief one task well and get a usable draft in one or two tries.
Reuse
Write your context down once as a skill, so you stop explaining yourself.
Delegate
Hand a whole repeating workflow to an agent, and keep the final say.
The steps you do yourself.
Let's talk
Tell us about your project or the problem you want solved. We reply within one business day.
- Greek or English
- A written proposal before any work starts
- Lead Developer at Valuedriven
- info@simple-ideas.gr
- +30 6947 004 708
Thanks, we got it. We'll reply within one business day.