JARVIS OS

J.A.R.V.I.S. — Just A Rather Very Intelligent System
Professional, offline-first AI Desktop Operating Assistant

---

Inspired by the Iron Man cinematic interface — JARVIS OS brings a sophisticated, voice-controlled, multimodal AI companion directly to your desktop.

Version: 1.0.0
Status: Production foundation (skills + voice + console)
License: MIT (see LICENSE)
Offline-first: Local by default (Ollama + formant/Piper TTS + SQLite)

---

🚀 Vision

A production-quality, secure, extensible desktop AI operating system that:

• Runs completely offline
• Understands natural language + voice + vision
• Automates complex workflows safely
• Learns and adapts over time
• Provides a beautiful, cinematic UI
• Is fully modular and plugin-driven

✨ Key Features

• Core AI Engine: Local LLM (Ollama) with tool-calling and long-term memory
• Voice Interface: Offline formant TTS with multiple male + female voices, optional Whisper STT / browser mic
• One-command skills: YouTube, Gmail, ChatGPT research, email drafts, timers, notes, maps…
• Mission planner: Multi-step commands (open youtube and set a timer…)
• Iron Man personality: Calm butler tone, always addresses you as “sir”
• Security: Sandboxed execution, permission system, audit logs
• Developer Friendly: Clean APIs, full test coverage, quality gate 10/10

---

🎙️ Voice profiles (choose any)

ID	Gender	Style
jarvis	Male	Classic deep British butler (default)
jarvis-fast	Male	Same character, quicker
deep	Male	Very low cinematic baritone
calm	Male	Slow, reassuring
warm	Male	Friendly mid-range
alert	Male	Urgent systems voice
news	Male	Clear broadcast style
aria	Female	Clear confident assistant
nova	Female	Bright energetic
friday	Female	Cool measured AI companion
soft	Female	Gentle lower tone
sage	Female	Warm professional narrator

In the console: pick a voice → Preview voice → send commands.
API: GET /api/v1/voice/voices · POST /api/v1/voice/speak with "voice": "aria"
Default: set JARVIS_UI_DEFAULT_VOICE=friday (or any id) in .env.

---

🏗️ Architecture Overview

JARVIS-AI/
├── agents/           # Orchestrator + mission planner + personality
├── backend/          # FastAPI (ai, voice, skills, memory, system)
├── core/             # Config + security (permissions, sandbox, audit)
├── skills/           # ★ Powers: browser, apps, email, ChatGPT, utils…
├── voice/            # Multi-voice formant TTS + voice service
├── frontend/public/  # Voice console → /console
├── docs/             # architecture · ROADMAP · QUALITY · assets (logo)
├── tests/ · scripts/
└── (next) electron/ vision/ plugins/

One-command examples (working now):

• Open YouTube / Open YouTube search lo-fi
• Open Gmail · open whatsapp
• ask chatgpt about machine learning
• help me write a cover letter
• email demo@example.com saying Hello from JARVIS
• open youtube and set a timer for 5 minutes (multi-step mission)
• System status · What time is it?

Setup guides: docs/guides/email_and_voice.md · .env.example
Roadmap: docs/ROADMAP.md · Architecture: docs/architecture.md
Quality: docs/QUALITY.md · Iron Man scope: docs/IRON_MAN_OS.md

---

📦 Tech Stack

Layer	Technology
Backend	Python 3.11, FastAPI, Uvicorn, Pydantic, SQLAlchemy
AI / LLM	Ollama (local)
Voice	jarvis-formant (multi-voice), optional piper-tts, faster-whisper
Memory	SQLite, contacts JSON
Frontend	Voice console (HTML/CSS/JS), React planned
Testing	pytest, quality_gate.py
Security	pydantic-settings, cryptography, sandboxing

---

🛠️ Quick Start (Development)

Prerequisites

• Python >= 3.11
• Git
• (Optional) Ollama + llama3.2 for deeper chat

1. Clone & Setup

git clone https://github.com/ARENJKY369/JARVIS-AI.git
cd JARVIS-AI

python -m venv .venv
source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -r requirements.txt

2. Start JARVIS

export PYTHONPATH=.
# Option A
python scripts/start_jarvis.sh

# Option B
uvicorn backend.app.main:app --host 127.0.0.1 --port 8000

Open the Voice Console:

http://127.0.0.1:8000/console — Main console (chat, TTS, skills)
http://127.0.0.1:8000/console_hud — Cinematic HUD (holographic AI core, voice, system panels)

1. Choose a Male or Female voice
2. Click Preview voice
3. Click ▶ Hear JARVIS or type a command

3. Run Full Validation

python scripts/quality_gate.py   # must be 10/10 all dimensions
python scripts/validate.py
pytest tests/unit/ -o addopts= -q

---

🎨 Branding & Assets

  

Asset	Path
Primary logo	docs/assets/jarvis_logo.png
Dark HUD logo	docs/assets/jarvis_logo_dark.png
App icon	docs/assets/jarvis_icon.png
Welcome audio	docs/assets/jarvis_welcome.wav

Colors: cyan #00E5FF · gold #C9A227 · void #05080F

---

🖥️ Console URLs

Interface	URL	Description
Voice Console	http://127.0.0.1:8000/console	Main voice + skills console (chat, TTS, skills)
HUD Console	http://127.0.0.1:8000/console_hud	Cinematic holographic AI core interface

Both served by the FastAPI backend. Open in browser after starting the server.

---

🔒 Security & Privacy

• All AI runs locally by default.
• No telemetry.
• Sandboxed automation.
• Explicit user consent for privileged operations.
• Full audit logging.

---

📜 License

MIT License. See LICENSE file.

---

JARVIS OS — "At your service, sir."