project-nomad

Product Introduction

1. Definition: Project NOMAD is a free, open-source, offline-first knowledge and education server software stack. Technically, it is a Docker-based, self-hosted server application that aggregates and serves curated open-source content and AI models locally.
2. Core Value Proposition: It exists to provide permanent, private, and resilient access to critical information and AI capabilities without an internet connection. Its primary value is in digital independence, emergency preparedness, and data sovereignty, offering a comprehensive offline knowledge base with an optional private AI assistant.

Main Features

1. Offline Information Library: Powered by Kiwix, this feature provides access to terabytes of static knowledge. It works by downloading and storing compressed ZIM files containing content like the full English Wikipedia, medical references from MedlinePlus, books from Project Gutenberg, and repair guides. The content is served via a local web server, allowing full-text search and browsing on any device connected to the NOMAD server's network.
2. Private, Local AI Assistant: Powered by Ollama, this optional feature enables private large language model (LLM) inference. It works by downloading open-source LLMs (like Llama 3.1 or Mistral) and running them entirely on the user's local hardware. All processing happens on the server's CPU or, preferably, a GPU (like an NVIDIA card), ensuring no data is sent to the cloud. Users can interact with the AI for chat, writing, coding, or document analysis with complete privacy.
3. Offline Vector Maps: Powered by OpenStreetMap data and served by a local tile server, this feature provides detailed, street-level world maps offline. It works by downloading regional map data (PBF files) for selected countries or the entire planet and rendering them dynamically in a web interface. Users can pan, zoom, and search locations without any internet connectivity, crucial for navigation in remote areas or during network outages.
4. Offline Education Platform: Powered by Kolibri, this module delivers structured educational content. It works by importing and serving learning resources like the full Khan Academy library (videos, exercises, articles) and other curated educational packages. It provides a classroom-like experience with user progress tracking, making it suitable for homeschooling or continuous learning in areas with unreliable internet.

Problems Solved

1. Pain Point: Critical dependency on fragile internet infrastructure for access to essential knowledge, educational materials, navigation, and modern AI tools. This creates vulnerability during power outages, natural disasters, in remote locations, or due to censorship.
2. Target Audience: Preppers and emergency preparedness advocates; off-grid living enthusiasts (homesteaders, RV/sailboat dwellers); technology enthusiasts and self-hosting advocates seeking data privacy; educators and families in areas with poor or expensive internet connectivity; researchers and professionals needing guaranteed access to reference materials.
3. Use Cases: Maintaining access to medical procedures and survival guides during a prolonged grid-down scenario; providing continuous education for children while traveling or living remotely; running a private AI coding assistant in a secure, air-gapped development environment; having reliable, detailed maps for navigation in the wilderness or on a boat; creating a permanent, personal digital library immune to link rot or content removal.

Unique Advantages

1. Differentiation: Unlike commercial competitors like PrepperDisk, Doom Box, R.E.A.D.I., or Z3R4H, Project NOMAD is completely free, fully open-source (Apache 2.0 license), and not locked to specific hardware like a Raspberry Pi. It supports multi-user network access and can leverage powerful, GPU-accelerated hardware for superior AI performance, whereas competitors often offer limited, static content on locked-down devices.
2. Key Innovation: Its modular, Docker-based architecture is its core technical innovation. This approach allows it to seamlessly integrate best-in-class open-source projects (Kiwix, Ollama, OpenStreetMap, Kolibri) into a single, cohesive system with a unified web interface. This "curated stack" model provides a turnkey experience that is more powerful and flexible than any single component or proprietary alternative.

Frequently Asked Questions (FAQ)

1. What are the hardware requirements for Project NOMAD? The recommended specifications for an optimal experience are an AMD Ryzen 7 with Radeon graphics or Intel i7+ processor, 32 GB of RAM, a 1 TB SSD, and a dedicated NVIDIA GPU for accelerated AI performance. It runs best on Ubuntu or Debian-based Linux, with support for Windows via WSL2.
2. Is Project NOMAD really free, and what's the catch? Yes, Project NOMAD is completely free and open-source software with no subscriptions or hidden fees. The "catch" is that you must provide your own hardware and technical effort to install and maintain it. The project is community-supported, with optional donations accepted to fund development.
3. Can I use Project NOMAD without the AI component? Absolutely. The AI assistant powered by Ollama is entirely optional. You can install and run Project NOMAD solely for its offline knowledge library, maps, and educational content. The system is modular, allowing you to enable or disable features based on your needs and hardware capabilities.
4. How do I update the offline content in Project NOMAD? Content updates are managed through the NOMAD web interface. While connected to the internet, the server can check for and download updated ZIM files for Wikipedia, new map regions, or refreshed educational packs from configured sources, ensuring your local knowledge base stays current.
5. How does Project NOMAD's AI privacy compare to ChatGPT? Unlike cloud-based AI like ChatGPT, Project NOMAD's AI runs 100% locally on your hardware. No prompts, data, or models are ever transmitted over the internet. This provides absolute privacy and security, making it suitable for processing sensitive information without risk of data leakage or corporate surveillance.