Muhammad Nouman Saeed

DevOps, data and AI engineer.

I build the infrastructure and the data pipelines, then the AI agents that run them.

Get in touch View LinkedIn

Arbisoft, Lahore
8+ years in production IT
5+ years in data and analytics

From fibre optics in 2009 to Kubernetes and AI agents today, my job has stayed the same: keep production running.

The stack

Six layers of the stack. One owner.

Scroll to take it apart.

Layer 1 · Cloud

AWS, Azure, Hetzner.

Scalable accounts and networks, secured and kept within budget. Proxmox on-prem.

Layer 2 · Infrastructure as code

Terraform and Ansible.

Every server and network written as code, reviewed and versioned.

Layer 3 · Containers

Kubernetes.

Self-managed clusters, AWS EKS and Hetzner Cloud, running in-house apps.

Layer 4 · Delivery

CI/CD and GitOps.

GitHub Actions and GitLab build it. ArgoCD keeps every cluster matching Git.

Layer 5 · Observability

Grafana, alerts in Slack.

Problems surface on a dashboard before anyone notices them.

Layer 6 · AI

Agents that run it.

Claude Code with MCP, Skills and sub-agents, n8n workflows and RAG on top of everything below.

AI agents, already in production.

Built with Claude Code and n8n, wired into the tools the team uses every day.

Slack assistant · RAG

Answers the team, from company knowledge.

# help-it ops-assistant

Teammate

How do I get VPN access for a new joiner?

ops-assistantAPP

Raise a "VPN access" ticket with their start date. IT approves within one working day.

it-handbook / access.md

Parallel agents · MCP

Ops work, running side by side.

Ticket triagesub-agent

Infra drift checksub-agent

Invoice syncn8n

Release notessub-agent

Claude Code
n8n
LangChain
Gemini

Data

AI analytics application

Designed and built alone, from the data pipeline to the insights.

Security

AI-assisted threat analysis

Anomaly detection built on Certified Ethical Hacker training.

Data, turned into decisions.

Data engineering, data science and data-driven products for organisations, automated from ingestion to the decision.

Retail chain · Power BI

700+ outlets, one screen.

Branch performanceAuto-refresh

Outlets700+

Countries2

ReportsDaily

Automated ingestion · ETL

Pipelines that load themselves.

WordPress
WooCommerce
Sheets

Looker Studioauto-refresh

Autonomous decisions · AI agents

From raw data to action, automatically.

IngestAPIs, stores, sheets
CleanETL and validation
ModelForecasts, anomalies
DecideAn agent, within rules
ActOrder, alert, report

Data science

Forecasts and anomaly detection on live BI data

Data products

Data-driven apps designed for the whole organisation

Automated reporting

Reports that build, refresh and send themselves

Natural-language queries

Ask the dashboard in plain English, over RAG

LLM reporting

Claude and Gemini write the summary under the chart

Sheets + Apps Script

An IT inventory app with no paid software

Power BI
Looker Studio
Tableau
BigQuery
Vertex AI
Python
Apps Script
Google Sheets
n8n

The numbers behind it.

8+years in production IT

700+retail shops synced nightly, Pakistan and Dubai

5+years of ETL pipelines, dashboards and reporting

7Odoo ERP modules rolled out

Career.

Jul 2022 – now
DevOps EngineerArbisoft, Lahore
AWS, Kubernetes, GitOps with ArgoCD, CI/CD, monitoring and AI automation.
Sep 2019 – Jun 2024
IT ManagerThe City School
Led the team that built the school's LMS; automation through COVID-19.
2018 – 2019
IT OfficerSefam
Nightly data from 700+ retail shops to one server.
2016
ICT OfficerTAC School System
Infrastructure and data warehousing across campuses.
2015 – 2017
IT OfficerDynamic Marketing Concepts
Branch IT in Karachi, Lahore and Islamabad.
2009 – 2010
Telecom InternPTCL
GPON systems and optical fibre.

Certified Ethical Hacker
AI & LLM Engineering, Coursera
Cloud & DevOps, Pluralsight
DevOps, Corvit
Agentic AI, Corvit
MSc IT
BS Telecom

Let's build something reliable.

Full-time roles and freelance projects in DevOps, cloud, data engineering and AI automation.

nouman4dears@gmail.com

View LinkedIn View GitHub