JARVIS area · for anyone who wants an AI assistant of their own

Your own Jarvis. Build your AI operating system, one piece at a time.

A Jarvis isn’t a chat. It’s an assistant that knows you, remembers what matters, uses tools and responds wherever you are, on Telegram, WhatsApp, or in the terminal. INEMA has a path to help you build your own: start with INTELECTO, where you choose the pieces and AI builds the code, then explore the Ready-to-install Jarvis assistants, the memory (a second brain), the courses and the videos. Everything is open and free.

Start with INTELECTO ↓ Find the right system ↓ View the courses ↓

Open projects and free courses. PT, EN, and ES where indicated.

01 · What is a Jarvis?

Every AI assistant is made up of the same seven pieces.

When you understand the pieces, you can choose better, spend less, and avoid getting locked into a product. They’re the same in INTELECTO, OpenClaw, Hermes, or a Jarvis you build yourself. The 7-piece video explains each one in 8 minutes.

1 · Channel

Where it talks

Telegram, WhatsApp, terminal, or voice. This is how you talk to it.

2 · Brain

Which model thinks

Claude, GPT, open models through OpenRouter, or local models through Ollama, without sending anything outside your device.

3 · Identity

Who it is

Text files with its personality, rules, and what it knows about you.

4 · Memory

What it remembers

A searchable local database that stores conversations, facts, and decisions.

5 · Tools

What it does

Read and write files, run commands, search the web, and connect to other systems.

6 · Loop

How it decides

It thinks, uses a tool, checks the result, and repeats until it’s done.

7 · Security

What it can’t do

A password vault, a list of forbidden commands, and approval before serious actions.

Result

An AI operating system

All seven pieces together, running on your computer or a server and working for you.

02 · INTELECTO · start here

You choose the pieces. AI builds your Jarvis.

The INTELECTO is INEMA’s way to build a personal assistant without bloated frameworks. It compares the leading open assistants, highlights what each one does well, and lets you put yours together like a grocery store: choose the ingredients, download a brief, and give it to Claude Code or Codex, which builds the code with tests.

Step	What you do	Where
1 · Understand	Watch the 7 components and see what each open assistant does well	Video of the 7 components
2 · Choose	At the parts shop, choose the channel, brain, memory, tools, and security	INTELECTO
3 · Request	Download the brief `meu-assistente.md` and hand it to your terminal agent	Claude Code or Codex
4 · Check	Run the tests and demo included with the code	On your computer
5 · Connect	Connect Telegram or WhatsApp and, if you like, deploy it to a 24/7 server	Installation video

Project · parts shop

INTELECTO

A comparison of open assistants, a parts shop, a build plan, and a sample Jarvis built from the brief, with tests.

Open → Course · PT · EN · ES

INTELECTO Course — from zero to expert

Six learning paths, from your first conversation with AI to the final path: your Jarvis.

Open the course → Workshop · PT · EN · ES

INTELECTO 4D — from zero to Jarvis

A two-day workshop to leave with your assistant built.

Open → Guide

INTELECTO — guide

The personal assistant in Python: Telegram, SQLite memory, and models through OpenRouter or Ollama.

Open the guide → WhatsApp

INTELECTO on WhatsApp

AI agent on WhatsApp via the Evolution API and OpenRouter, deployable with Docker.

Open → Video · PT

Install your Jarvis

From the parts shop to a live assistant, without writing code, in six minutes.

Watch →

Honestly. The brief alone isn't enough: it points to the INTELECTO build plan, and that plan is what gets the code right. The sample Jarvis runs and passes the tests on your computer; the Telegram channel needs your token, and WhatsApp and the 24/7 server are the next steps, explained in the video.

03 · Where to start

Say what you want. The table points you to the system.

The systems complement each other: one provides the assistant, another the memory, and another the tools. Choose the row that best matches your situation.

You want	The way	Use
Build your Jarvis from scratch	Choose each component and let AI write the code	INTELECTO
A complete local Jarvis	Notes, memories, decisions before the answer, Claude or Codex through your subscription	Jarvis v7
A model for programming your own	Base code in TypeScript, with action approval	Jarvis Modelo
Talk to it on Telegram	A queue that doesn't lose requests, cost per call, and local models	openpcbot v3 or inemaccbot
Talk to it on WhatsApp	An agent that runs in Docker	INTELECTO on WhatsApp
Use Claude Code or Codex as the foundation	A folder and layer structure that becomes your system	os-agentes or AIS-OS
Make it remember everything	A second brain in files, searchable by the agent	Astra 2Cérebro or Cérebro INEMA
Make it use any tool	A bridge for tools, permissions, and safeguards	INEMA Agent Runtime
Everything on my computer, with no cloud	Assistant with a local model, 100% private	Agente Hermes + Ollama
An assistant with a voice	You speak, it responds out loud	Assistente de Voz
Learn without programming	Short lessons, from zero to a ready-to-use assistant	Build your AI OS

04 · Ready-to-install Jarvis systems

Open systems with guides, ready to use and adapt.

Each one makes different choices about channel, brain, and security. Open the guide, run it on your computer, and swap out anything that doesn't work for you.

Jarvis · PT · EN · ES

Jarvis v7

Local personal assistant with notes, memories, five decisions before the answer (JEV Reflex), Claude or Codex through your subscription, and an image studio.

Open the guide → Model · PT · EN · ES