AbdullahShahid
I build systems that think. As an AI Automation Engineer at HostBreak, I've cut operational overhead by 60%, wired LLMs into live infrastructure, and shipped self-healing platforms that manage 10,000+ servers — without a human in the loop.
I'm drawn to the intersection of AI and real infrastructure: MCP servers, LLM quantization, n8n workflows, and Node.js backends that actually hold up under load. Currently exploring AI voice agents and autonomous server remediation.
Currently
Building: AI Voice Agent — real-time speech-to-action pipeline with LLM reasoning
Building: Automated Server Analysis Platform v2 — expanding to 50,000+ asset monitoring
Learning: Advanced LLM fine-tuning and RAG pipeline optimization
Currently
Building: AI Voice Agent — real-time speech-to-action pipeline with LLM reasoning
Building: Automated Server Analysis Platform v2 — expanding to 50,000+ asset monitoring
Learning: Advanced LLM fine-tuning and RAG pipeline optimization
About Me
I build systems that think. As an AI Automation Engineer at HostBreak, I've cut operational overhead by 60%, wired LLMs into live infrastructure, and shipped self-healing platforms that manage 10,000+ servers — without a human in the loop.
I'm drawn to the intersection of AI and real infrastructure: MCP servers, LLM quantization, n8n workflows, and Node.js backends that actually hold up under load. Currently exploring AI voice agents and autonomous server remediation.
Technologies I work with
- LangGraph / Multi-LLM
- FastAPI / Postgres
- n8n / UChat
- Node.js / Express
- Python
- Flutter / Dart
- MCP / LLM Ops
- Ansible / Linux
- Redis / Docker
- Git / GitHub
Education
BS Computer Science
Riphah International University
Available for new opportunities
Interactive CLI Playground
Execute live workflow simulations, inspect MCP server health, and trigger self-healing pipeline demos.
[INIT] Abdullah Shahid Interactive CLI Sandbox v2.4
Type "help" or click any shortcut button below to execute automated workflows.
Skills & Capability Matrix
Technical stack linked directly to verified real-world architecture implementations.
LangGraph
AI & LLMsStateful multi-agent candidate evaluation graph
LangChain
AI & LLMsDocument loading, prompt templates & RAG pipelines
Multi-LLM Router
AI & LLMsZero-downtime 6+ provider fallback & cost optimization
MCP Dev
AI & LLMsModel Context Protocol servers for DB & tool binding
LLM Quant.
AI & LLMsModel quantization (GGUF/AWQ) for low-latency local inference
Prompt Eng.
AI & LLMsStructured JSON output, tool calling & safety guardrails
Ollama
AI & LLMsLocal offline Llama 3 & DeepSeek model deployment
n8n
Automation10,000+ server self-healing workflows & reboot loops
UChat
AutomationConversational AI bot integration & channel routing
AI Workflows
AutomationEnd-to-end webhook automation & event-driven processing
Ansible EDA
AutomationEvent-driven Ansible rulebooks for autonomous server fixes
FastAPI
Full-StackHigh-throughput async Python APIs & background tasks
PostgreSQL
Full-StackStateful candidate relational DB & workflow tracking
Node.js
Full-StackScalable backend services & real-time webhook listeners
Express.js
Full-StackREST API endpoints & microservice integration
MongoDB
Full-StackDocument storage for unstructured payload logs & telemetry
Flutter/Dart
Full-StackCross-platform mobile UI for server status & monitoring
Python
Full-StackLangGraph agents, ML scripts & stealth CI auto-push
Redis
DevOps & InfraDistributed locking, rate-limiting & session caching
Ansible
DevOps & InfraInfrastructure as Code playbook automation for Linux fleets
Linux Admin
DevOps & InfraKernel tuning, systemd services & server administration
Git
DevOps & InfraStealth automated commit sync & Conventional Commits
WHM/cPanel
DevOps & InfraWeb hosting server administration & DNS automation
Click any skill card use-case box to view associated project architecture
Projects
Things I've built — from multi-agent AI systems to self-healing infrastructure.
HR_AI: Multi-Agent AI Hiring Platform & Automated DevOps Engine
Enterprise-grade candidate evaluation platform leveraging a stateful LangGraph multi-agent architecture, dynamic fallback routing across 6+ LLM backends (Gemini, OpenAI, Groq, Mistral, Cohere, Cerebras), and an automated stealth Git CI/CD sync pipeline.
JetBackup 5 Monitoring & Self-Healing System
Closed-loop backup remediation pipeline on the LOPA stack — detects failed/partial/missed backups across 5+ cPanel servers, auto-retries targeted accounts via JetBackup API, and publishes a live 5-tab Google Sheets dashboard. Zero human intervention required.
Automated Server Analysis Platform
Self-healing infrastructure platform that scans 10,000+ assets, auto-diagnoses issues across WordPress/Laravel/Node.js stacks, and remediates via SSH — no human needed.
Automated Bot System
WhatsApp + CRM chatbot that cut manual support work by 60% and response time by 90%, integrating n8n, UChat, WHMCS, and VTiger through a custom Node.js middleware.
AI Plant Diagnostics
End-to-end plant disease classifier — Python ML model, Flutter mobile app, Node.js/MongoDB backend, and a real-time Admin Panel connecting farmers with pathologists instantly.
Spotify Clone
Full-featured music streaming app with auth, playlists, and audio playback — Flutter frontend backed by a FastAPI/Python server.
Task Management App
Offline-first productivity app with task creation, categories, reminders, and progress tracking — all stored locally via SQLite.
Multi-LLM Outage & Fallback Simulator
Live interactive architecture playground demonstrating zero-downtime multi-provider LLM routing.
Test Zero-Downtime Multi-Provider Fallback Logic
Click provider badges below to simulate real-world API outages (503 / 429) and watch the router re-route traffic dynamically.
Live Routing Log Output:
Click "Simulate API Request" above to test dynamic fallback routing...
Fulfilled Provider
Awaiting routing run...
Experience
Where I've worked and what I've built.
AI Automation Engineer
HostBreak
Leading AI & infrastructure engineering — building self-healing systems, LLM integrations, and intelligent automations that manage 10,000+ production assets.
Key Contributions
- Architected MCP servers granting LLMs secure, real-time database access for agentic reasoning.
- Deployed GGUF/AWQ-quantized models on consumer hardware, cutting infrastructure costs significantly.
- Wrote Ansible playbooks for automated server provisioning — 100% environment consistency across AI and Node.js services.
- Built a NestJS/Next.js self-healing platform that auto-remediates 10,000+ assets via SSH job queues.
- Cut support response times by 40% with intelligent WhatsApp chatbot logic via UChat and n8n.
- Eliminated 60% of manual processing through Node.js middleware integrating WHMCS, WHM, and internal CRMs.
- Drove lead conversion with 100% real-time CRM tracking and automated multi-step follow-up sequences.
Quantified Automation ROI
Test the financial & operational impact of self-healing server infrastructure based on HostBreak benchmarks.
Quantify Server Automation Impact
Based on Abdullah's verified 60% operational overhead reduction benchmark at HostBreak.
Automated Tickets / Mo
9,000
60% L1 Support Reduction
Engineering Hours Saved
3,150 hrs
Per month saved
Estimated Annual Overhead Savings
$1,701,000 / year
Based on self-healing LLM & n8n workflow automation benchmarks
Get In Touch
Have a project in mind or want to collaborate? I'd love to hear from you.