The Maknoon Project is an educational initiative focused on designing and deploying a fully offline, autonomous AI system built from the ground up. Unlike cloud-based AI tools, this system runs entirely on local infrastructure using Ubuntu Server, Ollama for local large language model (LLM) capabilities, and OpenClaw for intelligent workflow orchestration and tool execution. The project demonstrates how modern AI systems can be engineered in a secure, private, and self-hosted environment while serving as a practical learning platform for AI infrastructure and automation.

The project is structured into three main stages:

Stage 1 – Building the AI Infrastructure:

Setting up a secure foundation by creating system backups, installing Ubuntu Server in a virtual machine, configuring Hyper-V, and deploying Ollama to run a local large language model.

Stage 2 – Enabling Local AI Interaction:

Installing and configuring Open WebUI to provide a browser-based interface for interacting with the local LLM in a controlled offline environment.

Stage 3 – Developing Autonomous Workflows:

Integrating OpenClaw to design intelligent, agent-driven workflows capable of executing structured tasks such as file analysis, presentation generation, or automated content creation.