Reshare — content from Eric Djavid
I just turned my Mac into a company of 7 AI employees.
It’s called JARVIS. And I’m giving the code away for free.
My problem: 7 Claude Code sessions open at the same time.
One per project. On 2 screens.
I spent my time looking for the right terminal.
And I was missing the ones that were waiting for me.
So I built my control station.
Concretely:
→ I switch terminals in a third of a second. One click, or “open number 3.”
→ Each session is a sphere on a 3D map. I am omniscient.
→ A master agent orchestrates the others. “Tell number 2 to run the tests.”
→ I talk to it like a colleague. In French.
→ It triggers actions, opens the dashboard of my apps.
Result: no more agents waiting on me without my knowing.
I run my 7 projects with it, from my Mac.
The code is open source. Installation takes 5 steps.
What you get:
The complete code of my version of JARVIS
The 3D map of all your Claude Code sessions
The French voice and the microphone that ignores its own voice
The guide to install it in 5 steps
FREE access! 
Reposts prioritized with absolute priority. 
LinkedIn Sources
- Profile: Pierre Jean Duvivier dit Sage (LinkedIn)
- Shared author: Eric Djavid (LinkedIn)
The post Je viens de transformer mon Mac en entreprise de 7 employés IA. appeared first on les plumes qui mordent.
Frequently Asked Questions
FAQs Turning Your Mac Into a Company of 7 AI Employees
1 What does it mean to turn my Mac into a company of 7 AI employees
It means running a set of AI agents or assistants on your Mac that each handle a different joblike writing research scheduling or customer repliesso they work together like a small team
2 Do I need a powerful Mac for this
Not necessarily Most AI agent setups run in the cloud and only need a browser or a lightweight app A local setup with opensource models needs more RAM and a decent chip but a standard modern Mac works for most cloudbased tools
3 What are AI employees exactlyare they real software
Theyre software agents powered by large language models Each one is configured with a role instructions and access to certain tools or files so it behaves like a specialized worker
4 How is this different from just using ChatGPT
Instead of one chat window you have multiple agents with separate roles that can hand off tasks share context and run on schedules Its more like a workflow than a single conversation
5 What kinds of jobs can the 7 AI employees do
Common roles include a writer a researcher an email assistant a scheduler a data analyst a customersupport responder and a project coordinator You can swap roles to fit your needs
6 What are the main benefits
Less repetitive work faster turnaround on routine tasks lower costs than hiring and the ability to run tasks 247 You also keep everything on hardware you already own
7 Do I need to know how to code
For cloudbased agent platforms nomost use a visual builder or simple settings Running local opensource agents usually requires some comfort with the terminal
8 How do the agents talk to each other
Through a shared workspace message queue or orchestration tool One agent finishes a task and passes the outputor a requestto the next agent in the chain
9 Whats the biggest mistake beginners make
Giving agents vague instructions or too much access at once Start with one clear role test it then add more agents gradually
10 How do I keep my data private
Use cloud tools with clear data policies or run local models
