I Build AI Automations,
Self-Hosted Infrastructure
& Intelligent Systems
Building autonomous AI systems from my private homelab using production-grade orchestration, secure self-hosted virtualization, and high-performance automated workflows.
Architecting the Homelab
Mohamed Tamer (3lmagary) is a self-taught automation architect, infrastructure engineer, and tech content creator. I design private clouds, segment localized container workflows, and construct offline RAG pipelines.
Engineering Philosophy
True digital sovereignty means owning your data and your hardware. I leverage virtualization clusters to self-host mission-critical databases, secure network tunnels, and automated agents. By building enterprise infrastructure in a homelab environment, I engineer systems that are autonomous, secure, and incredibly fast.
Through content creation, I share the blueprint of these systems—helping developers, tinkerers, and AI builders gain independence from centralized SaaS architectures.
observability-node-01
Core Domains & Current Focus
Key technical domains where I architect solutions, automate workloads, and deploy workflows.
AI Agents
Constructing multi-agent reasoning graphs, local LLM configurations (Ollama), and private retrieval-augmented generation (RAG) vector stores.
AI Automation
Bridging traditional APIs with modern LLM steps to automate complex processes, content generation, and code evaluation pipelines.
n8n Workflows
Designing advanced event-driven pipelines, webhook processing, API integrations, and robust automated workflows.
Self-Hosting
Building local backup nodes, secure private search engines, localized Git hosting, and sovereign storage platforms.
Infrastructure
Managing network bridges, container isolation, automated reverse proxies, cluster firewall configuration, and local DNS filtering.
Python Automation
Developing lightweight scripts, web-scrapers, file-system monitors, custom API adapters, and system management toolsets.
Linux Systems
Debian and Ubuntu server hardening, systemd automation configs, SSH key controls, user permission schemes, and shell scripting.
Active Projects in Development
Building an offline graph framework linking Llama3 nodes to query Wikipedia, ArXiv, and local files autonomously.
Writing a Python daemon to pull CPU temps, ZFS disk health, and RAM states to export metrics directly into Prometheus.
Automated transcoding cluster using FFmpeg in LXC container, triggered by webhooks when new video files land in storage.
Tech Stack & Infrastructure
Technologies and systems I interact with daily to architect and manage automation systems.
Python
Scripts & APIsn8n
Workflow EngineDocker
ContainerizationProxmox
VirtualizationLinux
OS KernelCloudflare
DNS & TunnelsAPIs
IntegrationsGitHub
Version ControlAI Tools
LLM IntegrationsOllama
Local LLM HostGrafana
Metrics DashboardTraefik
Reverse ProxyFeatured Systems & Projects
Production-grade systems running live inside my private infrastructure.
Sovereign AI Research Agent
An autonomous research agent running completely locally to prevent data leaks. Using n8n core loop and Ollama, the agent queries local document directories, processes PDF literature via vector searches, and generates formatted markdown digests.
[22:15:02] Loading vector index...
[22:15:04] Initializing Ollama: Llama-3-8B
[22:15:06] Executing multi-path web query...
[22:15:11] Context chunk retrieved. Score: 0.94
[22:15:13] Synthesizing final research report...
Self-Hosted Homelab Cluster
Bare-metal hypervisor cluster utilizing Proxmox VE for running containerized workloads, microservices, and databases. Configured with ZFS storage arrays for complete backup redundancy and isolated network bridges for secure operation.
Zero-Trust Reverse Proxy Gate
Automated request router and access portal. Uses Traefik to run Let's Encrypt certificates alongside Cloudflare Zero-Trust Tunnels, exposing resources securely without port forwarding. Local network blocks ads at DNS level via AdGuard Home.
[PROXY] API Request: app.3lmagary.com ➔ localhost:8080
[TLS] Let's Encrypt Certificate Valid. Renewal: 62d
[TUNNEL] Cloudflare Tunnel established safely via edge-01
[DNS] AdGuard Blocked: 1,402 queries blocked today
Infrastructure Metrics Collector
Centralized monitoring system collecting real-time CPU thermal loads, memory utilization, container lifecycle states, and network performance. Integrated metrics collectors (Prometheus) feed alerts directly to a visual Grafana cockpit.
Automated Deployment Workflows
Event-driven pipelines engineered to manage automated infrastructure configurations, build and run container clusters, back up vector databases, and execute custom Python scripts through standardized API endpoints.
{
"trigger": "webhook",
"workflow_id": "build_deploy_v2",
"steps": [
"pull_git_payload", "validate_hashes",
"deploy_container_stack", "send_status"
],
"status": "completed"
}
Technical Content Creator
Documenting my homelab journey, automation workflows, and AI implementations across social platforms.
YouTube
Detailed tutorials, homelab walkthroughs, and step-by-step n8n system guides.
Subscribe ChannelProfessional updates, systems design reports, and network automation insights.
Connect ProfileGitHub
Open-source repositories, automation templates, and custom exporter configurations.
Follow RepositoriesTwitter / X
Quick micro-updates, server snapshots, and automation tips directly from the lab.
Follow FeedSystems Engineering Journey
The educational path and milestones that led to advanced AI and infrastructure development.
Python Scripting Foundations
Learning syntax, building automation scripts, handling JSON files, interacting with web REST APIs, and writing local file-system scripts to automate repetitive tasks.
Linux System Hardening
Transitioning entirely to Linux. Managing headless servers, configuring user groups, learning bash shell commands, implementing firewalls, and managing systemd services.
Homelab Infrastructure & Clusters
Building local virtual structures using Proxmox hypervisor. Provisioning LXC containers, building virtual bridge networks, and setting up centralized storage stacks.
n8n Pipeline Automations
Architecting workflow logic. Linking webhooks, connecting calendars, scraping APIs, and writing automated pipelines that process events in real time.
Intelligent AI Systems
Integrating APIs with Large Language Models. Designing prompts, managing memory variables, and automating workflow actions with generative capabilities.
Autonomous AI Agent Networks
Deploying localized neural models (Ollama). Linking models to RAG structures (ChromaDB) to build sovereign agents acting on private data arrays.
Let's build sovereign infrastructure.
I build autonomous workflows and private homelab clouds. If you want to automate processes, set up self-hosted environments, or connect LLMs to databases, let's talk.