Local AI | Agents | Automation
Your own AI team
Eight short audio steps on building a fleet of local AI agents that do real work, read your inbox, run your shop, and heal themselves.
Most people use AI as a tool they pick up and put down. I run a team of agents that quietly own real parts of my work, all day, on my own machine.
This is the next level after the local AI basics. Not one assistant, but a designed team, with an orchestrator, a shared brain, guardrails, and the ability to actually act, all kept safely in check by a human.
It is the thing I genuinely do every day, and it has given me my hours back. This is how to build it without it ever doing something it should not.
Who it is for
- ·Founders and makers who already use local AI and want to go from one assistant to a real team.
- ·People drowning in repetitive work who want agents to own it, safely.
- ·Anyone curious how an autonomous agent fleet actually works in practice, from someone who runs one.
What you get
·Eight short audio lessons, each with a written transcript.
·How to design a team of focused agents and an orchestrator that routes work.
·A shared brain so your agents know you, and guardrails so they never overstep.
·A self-healing setup, and the one rule that keeps a powerful fleet safe.
The steps
8 steps
- 01
From tools to teammates
The shift that makes everything possible: stop using AI as a tool you pick up, and start managing teammates who own an area and run continuously.
€4
Preview →
- 02
Designing your team
Design before you build. One agent, one clear job, with a written spec of what it owns, what it must never do, and when to ask a human.
€4
Preview →
- 03
The orchestrator
The lead agent that turns a pile of agents into a team, taking incoming work and routing it to the right teammate so you stop being the switchboard.
€4
Preview →
- 04
A shared brain
A common memory every agent draws from, your rules, your voice, your history, so the team sounds like it came from inside your world and gets wiser over time.
€4
Preview →
- 05
Letting agents actually do things
Crossing the line from talking to acting. Tools, and the crucial split between reading, reversible doing, and the irreversible actions that need the heaviest care.
€4
Preview →
- 06
Guardrails and the human switch
The safety system that lets you sleep: an approval queue for anything irreversible, hard walls that cannot be crossed, and a kill switch you build before you need it.
€4
Preview →
- 07
A self-healing fleet
Agents break constantly. Watchdogs and heartbeats that restart them, and the circuit breaker that stops a doom loop from turning a small failure into a disaster.
€4
Preview →
- 08
Living with a team that works for you
What it is actually like day to day: managing rather than coding, earning trust slowly, keeping the human in the loop, and keeping the whole thing in service of your life.
€4
Preview →
The whole workshop
€29 bundle
Saves €3 vs buying each step.
Questions
- Do I need to be technical?
- Some comfort with your machine helps, but the real skill here is management and delegation, not coding wizardry. It is closer to leading a small team than to programming.
- Is this safe? I do not want agents emailing my customers.
- That is the heart of it. A whole step is guardrails, the approval queue, hard walls, and the kill switch, so your agents prepare everything and you approve anything that touches the real world.
- Can I skip around?
- Yes, though this one builds on itself more than most. The order is the path I would actually walk a friend through, from one agent to a whole fleet.
Want a workshop on a topic I have not added yet?
Write to me →