marziel 1.0.1

Marziel OS — Private AI Operating System with adaptive inference, 3-tier memory, and process management

🧠 Marziel OS — v1.0.1

Private AI Operating System — runs entirely on your hardware.

One command to boot your own AI OS — with MarzielFlow adaptive inference, 3-tier encrypted memory, and autonomous process management.

Marziel OS is a private AI operating system that transforms your machine into an intelligent, self-managing system. It features an AI kernel with persistent event loop, 3-tier memory (working → long-term → episodic), Unix-like process management, MarzielFlow adaptive inference optimization (T/Z distribution quantization, fuzzy logic, attention sink), TurboQuant KV cache compression, real-time maritime intelligence, and 12-route semantic intent routing — all without sending a single byte to the cloud.

Quick Install

pip install marziel
marziel serve

That's it. The right engine is auto-installed for your hardware. The first run downloads the model from Hugging Face. Everything is automatic.

What's New in v1.0.1

🧠 Marziel OS — The AI Kernel

Marziel is no longer just a chatbot. It's now a persistent operating system with:

  ╔══════════════════════════════════════════════════════════╗
  ║  🧠 Marziel OS v1.0 — Private Intelligence OS          ║
  ╚══════════════════════════════════════════════════════════╝

     Kernel:     active (heartbeat: 60s)
     Engine:     marziel-v6 (CUDA, 52.9 tok/s)
     Flow:       5 components (T/Z + ABQ + Fuzzy + Sink + Spec)
     Memory:     working=0 | long-term=2.4KB | episodic=156KB
     Processes:  0 running, 0 queued
     Security:   Sentinel active (anti-fingerprint, encrypted)

🔬 MarzielFlow — 5-Component Adaptive Inference

Component	What It Does	Impact
Speculative Decoding	Prompt lookup draft/verify	Higher throughput
T/Z Distribution Quantization	Per-layer distribution detection + Lloyd-Max boundaries	Better quality at lower bits
Adaptive Bit-Width (ABQ)	2/3/4-bit per layer by sensitivity	2.88-bit avg, memory savings
Fuzzy Logic Controller	Mamdani-style adaptive resource control	Smooth parameter adaptation
Attention Sink Cache	StreamingLLM: 4 sink tokens + 1024 rolling window	75% memory savings

🖥️ OS API

GET  /os/status       — Full kernel + engine + memory status
GET  /os/memory       — 3-tier memory (working, long-term, episodic)
GET  /os/ps           — Process list (like Unix ps)
GET  /os/top          — Real-time resource monitor
POST /os/recall       — Semantic memory recall
POST /os/remember     — Store encrypted persistent memory
POST /os/schedule     — Schedule recurring tasks
POST /os/kill/:pid    — Kill a process

Features

Feature	Description
🧠 AI OS Kernel	Persistent event loop, heartbeat, task scheduler, signal bus
💾 3-Tier Memory	Working (volatile) → Long-Term (AES-256 encrypted) → Episodic (permanent summaries)
⚙️ Process Manager	Unix-like ps, top, kill with priority scheduling
🔬 MarzielFlow	5-component adaptive inference optimization pipeline
⚡ TurboQuant	PolarQuant + QJL 3-bit KV cache compression (3x memory reduction)
🧠 AI Chat	Conversational AI with personality, code generation, data analysis
📡 Maritime Intelligence	Real-time GDACS, GDELT, NASA EONET, AIS, market data, bunker prices
💰 Market Data	Live Brent crude, BDI, EU Carbon, freight rates, forex, 12-port bunker prices
🚢 Vessel Integration	AIS tracking, fleet management, voyage data, maintenance records
🔌 Third-Party Connectors	MarineTraffic, VesselFinder, Spire, Datalastic, or any REST API
🔍 Web Search	Live internet results via DuckDuckGo — always up-to-date answers
🌐 Browser Control	Headless Chromium automation — navigate, click, type, screenshot
🔗 URL Analysis	Share any link — reads, extracts, and analyzes content
🐙 GitHub Analysis	Paste any repo URL — get architecture, languages, README analysis
⚡ Autonomous Agent	Multi-step reasoning with bash, file, web, browser, and maritime tools
📑 Hybrid RAG	Upload documents → reasoning + vector search → answers with citations
🔒 100% Private	Everything runs locally. Zero cloud dependency
🛡️ Sentinel Security	Anti-fingerprinting, encrypted storage, secure memory wipe, zero telemetry

Architecture

┌─────────────────────────────────────────────────────────────┐
│                      MARZIEL OS v1.0                        │
│                                                             │
│  ┌──────────────────────────────────────────────────────┐   │
│  │ INTENT SHELL (12 semantic routes)                    │   │
│  │ "check vessels" → parse → schedule → report          │   │
│  └──────────────────────────────────────────────────────┘   │
│                           │                                 │
│  ┌────────────┐  ┌────────▼───────┐  ┌──────────────────┐  │
│  │ PROCESS    │  │ KERNEL         │  │ MEMORY MANAGER   │  │
│  │ MANAGER    │  │                │  │                  │  │
│  │            │  │ Event Loop     │  │ Working Memory   │  │
│  │ Task Queue │◄─┤ Scheduler      ├─►│ Long-Term (AES)  │  │
│  │ PID Table  │  │ Signal Bus     │  │ Episodic (JSONL) │  │
│  │ Lifecycle  │  │ Heartbeat      │  │ Semantic Recall  │  │
│  └────────────┘  └────────────────┘  └──────────────────┘  │
│                           │                                 │
│  ┌────────────────────────▼─────────────────────────────┐  │
│  │ INFERENCE ENGINE                                      │  │
│  │ MarzielEngine + MarzielFlow + TurboQuant              │  │
│  │ ┌──────────────────────────────────────────────────┐  │  │
│  │ │ T/Z Quant │ ABQ │ Fuzzy │ Sink Cache │ Spec Dec │  │  │
│  │ └──────────────────────────────────────────────────┘  │  │
│  └───────────────────────────────────────────────────────┘  │
│                           │                                 │
│  ┌────────────────────────▼─────────────────────────────┐  │
│  │ NEURONS (Axon + Dendrite + 8 Synapse types)           │  │
│  │ SENTINEL (Security: anti-fingerprint, encrypted vault) │  │
│  └───────────────────────────────────────────────────────┘  │
│                    YOUR MACHINE                             │
└─────────────────────────────────────────────────────────────┘

Maritime Intelligence (Enterprise)

Marziel integrates real-time maritime data from multiple global sources:

┌───────────────────────────────────────────────────────────┐
│                   MARZIEL MARITIME INTEL                   │
├───────────────────────────────────────────────────────────┤
│  OSINT (Free APIs)         │  Platform Integration        │
│  ─────────────────         │  ────────────────────        │
│  📡 GDACS — Disasters      │  🚢 Fleet Vessels            │
│  📰 GDELT — Geopolitics    │  🗺️  Active Voyages          │
│  🌍 NASA EONET — Climate   │  ⚠️  Risk Zones              │
│  💰 Yahoo Finance — Market │  🏗️  Maintenance             │
│  💱 ExchangeRate — Forex   │  📋 Incidents                │
│  🚢 AISStream — Tracking   │  ⛽ Fuel Records             │
│                            │  🏭 Port Directory           │
├───────────────────────────────────────────────────────────┤
│  Third-Party Connectors (Enterprise)                      │
│  ─────────────────────────────────                        │
│  MarineTraffic · VesselFinder · Spire Maritime            │
│  Datalastic · OpenWeather · Custom REST API               │
└───────────────────────────────────────────────────────────┘

Maritime API Endpoints

GET  /api/maritime/report           — Full intelligence report
GET  /api/maritime/data/market      — Live market data
GET  /api/maritime/data/bunker      — 12-port bunker prices
GET  /api/maritime/data/fx          — Exchange rates
GET  /api/maritime/data/gdacs       — Disaster events
GET  /api/maritime/data/gdelt       — Geopolitical news
GET  /api/maritime/data/nasa        — Natural events
GET  /api/maritime/data/ais         — Vessel tracking
GET  /api/maritime/data/vessels     — Fleet data
GET  /api/maritime/data/voyages     — Voyage data
GET  /api/maritime/data/risk-zones  — Active risk zones
GET  /api/maritime/data/ports       — Port directory
GET  /api/maritime/connectors       — List connectors
POST /api/maritime/connectors/query — Query third-party services

Requirements

• RAM: 8GB minimum (16GB recommended)
• Disk: ~5GB for the model
• Python: 3.10–3.14
• Supported platforms:

Platform	Engine	Model	Auto-installed
Apple Silicon (M1–M4)	MLX	MLX 4-bit (4.5GB)	mlx-lm
NVIDIA GPU (any)	Marziel Engine	GGUF Q4_K_M (4.8GB)	llama-cpp-python
CPU (any x86/ARM)	llama.cpp	GGUF Q4_K_M (4.8GB)	llama-cpp-python

• Linux: Ubuntu, Debian, Arch, Fedora, RHEL, Alpine — any distro with glibc
• macOS: Intel (llama.cpp) or Apple Silicon (MLX)
• Windows: WSL2 recommended

Usage

# Install and boot
pip install marziel
marziel serve

# Check OS status
curl http://localhost:8001/os/status

# Store a memory
curl -X POST http://localhost:8001/os/remember \
  -H "Content-Type: application/json" \
  -d '{"key": "preference", "value": "maritime focus", "tier": "long_term"}'

# Recall a memory
curl -X POST http://localhost:8001/os/recall \
  -H "Content-Type: application/json" \
  -d '{"query": "preference", "tier": "auto"}'

# Schedule a recurring task
curl -X POST http://localhost:8001/os/schedule \
  -H "Content-Type: application/json" \
  -d '{"name": "vessel_monitor", "interval_seconds": 300}'

# Process management
curl http://localhost:8001/os/ps      # List processes
curl http://localhost:8001/os/top     # Resource monitor

Then open http://localhost:8001 — your private AI OS dashboard.

Pricing

Plan	Price	Features
Free	€0/mo	AI Chat, Web Search, URL Analysis, GitHub Analysis, OS Status
Pro	€20/mo	+ Agent Mode, Browser Control, Process Management, Scheduling
Enterprise	€199/mo	+ Maritime Intelligence, Connectors, Reports, Custom Fine-tuning

Links

• 🌐 Website: marziel.com
• 💻 GitHub: github.com/efops/marziel
• 🤗 Model: huggingface.co/efops/marziel-8b-custom
• 📦 PyPI: pypi.org/project/marziel

Contact

📧 [e-mail]

License

MIT License

---

Built with ❤️ by Efe (Efkan Isazade)