Setting Up an Offline Learning Environment
Project NOMAD allows users to create a self-contained knowledge server without requiring an internet connection after installation. This setup is particularly useful for educational institutions, remote locations, or users who need reliable access to information without relying on external networks. The installation process is terminal-based and supports Debian-based systems, with a recommended setup on Ubuntu 26.04 LTS. Once installed, users can access all features through a browser, eliminating the need for a desktop environment. The system is designed to be lightweight, with a minimal requirement of 5 GB of disk space and 4 GB of RAM for the base installation, making it suitable for a wide range of hardware configurations.
Integrating AI and Data Tools Locally
NOMAD includes an AI Assistant that supports local chat with document upload and semantic search using Qdrant. This feature allows users to interact with a knowledge base without needing an internet connection. The AI can be powered by Ollama or OpenAI-compatible tools like LM Studio or llama.cpp. For data management, CyberChef provides encryption, encoding, and analysis tools, all accessible through the browser. These tools are containerized and managed via Docker, ensuring that users can run complex applications without manual configuration. The system also supports running custom Docker containers, offering flexibility for advanced users who want to integrate their own applications.
Managing Educational Content and Resources
The education platform in NOMAD is built on Kolibri, offering access to Khan Academy courses with progress tracking. This makes it ideal for students and educators who need offline learning materials. Additionally, the Information Library includes offline Wikipedia, medical references, and ebooks via Kiwix, ensuring that users have access to a vast amount of information without an internet connection. Users can also download regional maps through ProtoMaps for offline use, making it a comprehensive solution for both educational and navigational needs. The Supply Depot provides a one-click app catalog, simplifying the process of accessing additional tools like PDF readers, file browsers, and password managers.
Optimal Hardware and System Requirements
While NOMAD itself is lightweight, the performance of AI tools and other resource-intensive features depends on the hardware. For a basic installation, a 2 GHz dual-core processor, 4 GB of RAM, and 5 GB of storage are sufficient. However, for running AI models like Ollama, the recommended specifications include a 32 GB RAM system, an NVIDIA RTX 3060 or equivalent GPU, and at least 250 GB of storage. These requirements ensure that users can leverage the full potential of the AI Assistant and other advanced features. The system is designed to be hardware-agnostic, allowing users to choose components based on their specific needs and budget.
Controlling Network and Privacy Settings
Project NOMAD is designed for offline use, requiring an internet connection only during installation and when downloading additional resources. The system includes a connectivity test that uses Cloudflare's endpoint by default, but users can customize this through the UI or environment variables. This flexibility ensures that the server can operate in various network environments, including those with restricted access. Additionally, NOMAD does not include any telemetry, prioritizing user privacy. Users can control which services are exposed on the network by managing port settings, providing an additional layer of security for shared environments.
Project NOMAD is ideal for users who need reliable offline access to educational resources, AI tools, and data management features, with a setup that balances flexibility and performance.