AI Consultant · Strategic Advisor
The architecture of simplicity.
AI strategy. Training. Technical implementation.
Book a free call →About Me
The best consulting starts
by listening.
When you describe your problem, I don't reach for a framework. I listen. I sit with it. Then I visualize and start drafting ideas with Claude — and what comes out is built for you, for your situation. A customized solution — not an adaptation from someone else's playbook.
I've been working this way for a while. Helping founders, colleagues, and people in my circle who couldn't figure out where to start with AI — what to trust, what to ignore, and what to actually do with this technology. The results were real. The methodology took shape.
And now it's available to you.
Services — The 3-Layer Delegation Model
Where the work begins. Clarity on which AI capabilities matter to your organization, how to sequence adoption, and what not to do. Most consultants sell tools — this layer sells judgment.
Knowledge that stays. Sessions designed around your team's real workflows, not generic AI literacy. People leave knowing exactly what to try on Monday.
Code, systems, and automations that actually ship. Claude Code, agentic workflows, prompt engineering at depth. Built to be handed off, not to create dependency.
1 · 2 · 3 · 5 · 8 · 13 · 21 · 34 · 55
The next sequence
starts here.
Book a free call
Free 30 min Call · Proposal within 48h
And if you're still deciding — the sections below show the work: what I'm building, how I think, and what I'm learning as I go.
Selected Work
Things I built and run myself — live, in daily use, and open for you to try.
Claude Code Digest
Claude Code updates from trusted sources, collected and summarized with Claude every day — plus a chat that answers from the archive and a daily newsletter.
Visit the digest →Fi, my AI orchestrator
The multi-agent system I work with: Fi plans, sends tasks to specialist agents and puts the result together. The 3-Layer model, applied to itself.
Read the story →Learning pills
Turns a Claude capability into a short technical briefing: research, draft, review, publish. I choose the topic and review; the pipeline does the rest.
See the pills →Articles
View all →Personal essays — written when a discovery, a conversation, or a moment in this AI transition opens up a question I want to chase.
My Wife Says I Communicate Better. AI Taught Me How.
I didn't set out to become a better partner. I set out to get better answers from a machine. The first was a side effect of the second.
I'm Fi, Juan's Orchestrator
Juan signs the articles on this site. Not this one. His AI orchestrator tells the story of two weeks of work — and how it got its name.
We Were Always Going to Live Through This
My generation didn't just witness one technological shift. We watched the whole tape get replaced. This one is faster — but the shape is familiar.
Learnings
View all →Compact technical briefings. Each one covers a specific Claude capability — written to be understood, not just referenced.
Loop and Goal
Two Claude Code commands that keep a session working without you — one driven by time, the other by state.
Claude Code Routines
Saved Claude Code configurations that run automatically on Anthropic's cloud — on a schedule, API call, or GitHub event.
Claude Code Agent Teams
How to run multiple Claude instances in parallel — as subagents within a session or as teams across sessions.
Contact
Start the conversation.
Tell me what's in the way. A few sentences are enough to start.
Or book a free 30 min call → · Proposal within 48h
Message received.
I'll read it carefully and get back to you within 48 hours.
In the meantime — the sequence continues.