Available for work — Lahore, PK

Muhammad Umair

Cyber Security / Python / AI Automation

I design and build secure systems, AI-driven automation and full-stack products — from AI agents that handle real business conversations to algorithmic trading bots and offline-first retail tools. Currently researching Cyber Defense & Threat Intelligence.

Start a project → Resume / CV ↓ Selected work GitHub LinkedIn
01 Security 02 AI & Automation 03 Full-Stack
Portrait of Muhammad Umair
BS Cyber Security · UCP LAHORE
Based Lahore, Pakistan Focus Security · AI · Automation Response < 24 hours Scroll
01 — About

Secure by default,
intelligent by design.

Building secure systems and AI-driven automation — from enterprise security workflows to agents that quietly run in the background.

I'm Muhammad Umair — a Cyber Security Specialist, Python Developer and AI Automation Engineer based in Lahore, Pakistan.

Currently researching Cyber Defense & Threat Intelligence. I architect secure enterprise solutions, build AI-driven automation pipelines, and ship full-stack web experiences where performance and reliability are non-negotiable.

Pursuing a BS in Cyber Security at the University of Central Punjab (UCP) after completing ICS Physics at Punjab Group of Colleges. I'm a Microsoft Learn Student Ambassador (Community Influencer track, ID: studentamb_625350) and currently learning Advanced Penetration Testing and Python Automation.

My work spans offensive security, AI agent development, algorithmic trading systems, and full-stack web architectures. I care about the things most people skip: threat modelling before the first commit, observability before launch, and documentation that someone can actually read at 2am.

Based in
Lahore, Pakistan
Education
BS Cyber Security · UCP
Prior
ICS Physics · Punjab Group
Focus
Cyber Defense · AI · Automation
Stack
Python · JS · SQL · MQL5
Ambassador
MLSA Community Influencer
Availability
Open to opportunities
0 Projects Shipped
0 Certifications
0K+ Organic Views
0 Years Building
02 — Capabilities

Technical Focus.

Six disciplines, one mindset: understand the system, find the weak point, then build it better.

01

Offensive & Defensive Security

Vulnerability assessment, threat modelling and incident response workflows — practiced through hands-on labs and enterprise-style security exercises.

Kali LinuxAuditingThreat IntelIR
02

AI Agents & Automation

Conversational agents and automated pipelines that remove manual work — wired into real business processes, not demos. Built with AI-assisted workflows using n8n and Ollama.

GeminiClaudeOllaman8n
03

Python & Backend Engineering

Clean, testable Python — from automation scripts and FastAPI services to data pipelines that have to run unattended. Containerized with Docker.

PythonFastAPIDockerBash
04

Algorithmic Trading

Expert Advisors and strategy systems for MetaTrader 5, with strict position sizing and drawdown controls baked into every build.

MQL5MetaTrader 5Risk
05

Full-Stack & 3D Web

Fast, accessible front-ends and the back-ends behind them — including WebGL and Three.js experiences that don't melt laptops.

JavaScriptThree.jsWebGLHTML5
06

Data, Cloud & Growth

SQL data modelling, cloud form pipelines with Tally.so and analytics — plus SEO work that has driven six-figure organic reach.

SQLTally.soSEOAnalytics
03 — Production Engineering

Selected Engineering Systems.

Hardened security utilities, autonomous compliance agents, real-time telemetry, conversational AI, and full-stack systems.

Project 01 — Cyber Intelligence

VIXOR — Autonomous Cyber Intelligence & Attack Surface Analyzer

Multi-engine vulnerability scanning orchestration platform that aggregates Nuclei, Nmap, and ZAP results, prioritizes risk using machine learning, and generates explainable remediation advice (SHAP) for SOC teams. Containerized with Docker and served via FastAPI.

PythonNucleiNmap ZAPFastAPISHAP (Explainable AI) Docker
Explore Repository
vixor_scan_dashboard.png
Scan Engines3 Nuclei · Nmap · ZAP
ExplainabilitySHAP per-finding
Project 02 — Endpoint EDR

auto-incident-responder — Linux Endpoint EDR

Open-source active defense containment daemon for Linux production environments. Monitors process lifecycles, privilege escalations, and abnormal sockets via eBPF / kernel hooks. Executes targeted PID termination, volatile memory dumps, and iptables isolation before human IR teams can respond. Includes an interactive telemetry terminal.

PythoneBPFLinux Kernel Hooks FastAPIDockerWebSockets
Explore Repository
edr_containment_terminal.png
auto-incident-responder Dashboard
SYSTEM ONLINE MEM 22% MONITORING
* Client-side deterministic simulation — No live socket execution
Project 03 — Compliance Auditor

Ghost Data Auditor — GDPR Deletion Verification Agent

Verifies post-deletion user data across logs, cache layers, cold storage, and search indexes to eliminate ghost compliance liabilities. Automated audit pipelines with NVIDIA NIM acceleration, containerized with Docker and served via FastAPI.

NVIDIA NIMPythonFastAPI DockerAutomated Audit PipelinesGDPR
Explore Repository
ghost_audit_pipeline.png
Ghost Data Auditor Dashboard
Scan TargetsLogs · Cache · Cold · Index
ComplianceGDPR Verified
DiscoveryScan Layer NIMVerify CertifyDeletion RemediateGhost Data ReportAudit Trail
Project 04 — AI Business Agent

LocalBiz AI WhatsApp Agent

Autonomous 24/7 conversational receptionist for regional businesses, integrated with Meta Cloud API. Engineered with dual-engine failover (Groq Llama 3.3 primary → Gemini 2.0 backup). Validates table bookings, qualifies buyer intent, and pushes clean lead payloads into staff channels.

FastAPIMeta Cloud APIGroq Llama 3.3 Gemini 2.0 FailoverDocker
Explore Repository
whatsapp_agent_session.png
LocalBiz AI WhatsApp Agent live interface
Uptime LIVE99.8% SLA
Avg Response1.2s P95
agent▸
ClientWhatsApp FastAPIGateway Groq Llama3.3 · Primary Gemini 2.0Fallback OrderDispatch
routing ▸ Groq Llama 3.3 (primary)
Project 05 — Multimodal GenAI Studio

HIGGSFIELD-AI-OPEN-SOURCE

Multi-provider BYOK (Bring-Your-Own-Key) architecture for generative image and video synthesis. Routes prompts across NVIDIA NIM, Replicate, and Fal API through a zero-logging proxy. Built with Next.js / Vanilla JS and containerized with Docker.

Next.jsVanilla JSNVIDIA NIM ReplicateFal APIZero-Logging Proxy Docker
Explore Repository
higgsfield_genai_studio.png
HIGGSFIELD GenAI Studio
ProvidersNIM · Replicate · Fal
LoggingZero BYOK
Project 06 — Sentiment Telemetry

BrandPulse AI Telemetry

Automates sentiment audit and brand protection for local food chains and retail outlets. Connects scattered review sources across Google Maps and social portals, parses nuance with local context awareness, and drafts humanized resolution responses before customer dissatisfaction impacts public rating thresholds.

PythonFastAPISentiment NLP Brand DefenseAutomated Feedback
Explore Repository
brandpulse_dashboard.png
BrandPulse sentiment analytics dashboard
Project 07 — Offline-First POS

RetailFlow-POS (Offline-First)

Brick-and-mortar point-of-sale for environments prone to sudden internet outages. Zero subscription fees, zero external server dependencies. Runs completely client-side with instant billing, ledger credit-book tracking (Khata), inventory reconciliation, and direct-to-WhatsApp digital receipt dispatching.

HTML5Vanilla JavaScript LocalStorage EngineOffline-FirstWhatsApp Receipt Dispatch
Explore Repository
retailflow_checkout_ui.png
RetailFlow-POS checkout console
Offline Mode∞ ACTIVE
Receipts2.4K SENT
Project 08 — 3D Storefront

Bella's Fast Food — 3D Storefront & Reservation Engine

High-conversion digital storefront with a Three.js / WebGL 3D experience, dynamic table layout engine, and one-tap WhatsApp order routing. Built for real-world restaurant operations with reservation and order management.

Three.jsWebGLDynamic Table Layout WhatsApp Order Routing
bellas_3d_storefront.png
Bella's Fast Food Haripur 3D digital storefront
MENU
TABLE
ORDER
RESERVE
PAY
WHATSAPP
Project 09 — Quantitative Execution

Algorithmic Risk-Capped Trading Bots

Automated Expert Advisors (EAs) and quantitative execution monitors for MetaTrader 5 with algorithmic risk governance: hard-coded 2% drawdown limits, dynamic position sizing, and slippage filters. Phase-1 architecture successfully validated, completed, and shipped/gifted to a fellow developer.

MetaTrader 5MQL5 PythonQuantitative RiskTransferred / Shipped
Architecture Complete · Gifted to Developer
tradingbotpic.jpg
Algorithmic Risk-Capped Trading Bot Dashboard
Project 10 — Data-Driven Distribution

Sports Analytics & Organic Distribution Engine

Diagnosed severe indexing stagnation on a sports entertainment publication. Restructured URL taxonomy, rectified internal crawl budget waste, and rewrote programmatic metadata schemas to match high-velocity search queries during live fixtures. Catapulted impressions from dormant levels to 230K+ verified organic views in 7 days without ad spend.

Search ArchitectureMetadata Engineering Crawl OptimizationOrganic Growth Strategy
Verified Growth Case Study (230K+ Views)
search_console_performance_spike.png
Google Search Console 230K organic impressions spike
04 — Recognition

Credentials & Recognitions.

Certifications, competitions and programmes that shaped how I build.

Aug 2026

Alibaba Cloud AI Hackathon Pakistan

Grade 2A project placement under the Bano Qabil platform, architecting an AI-driven solution end to end.AI Cloud

Mar 2026

International Research Olympiad

Top 20% global ranking — advanced analytical and research problem-solving under competitive conditions.Global Recognition

2026

Microsoft Applied Skills — Cloud Security & Monitoring

Validated expertise in securing cloud environments and deploying comprehensive monitoring tasks.Cloud Security

2026

Google Foundations of Cybersecurity

Coursera certification covering threat intelligence, network security protocols and risk mitigation strategy.Foundations

2026

Deloitte Cybersecurity Job Simulation

Hands-on enterprise incident response and infrastructure security analysis workflows.Enterprise

2026

Microsoft Learn Student Ambassador

Community Influencer track (ID: studentamb_625350) — bridging student communities and enterprise tech.Ambassador

2026

Major League Hacking (MLH) Fellowship

Fellowship applicant — featured open-source security tooling, endpoint containment systems, and scheduled automation workflows.Open Source

2026

Microsoft Learn — SQL & Data Security

Specialised training on SQL database objects and strict data security compliance.SQL & Security

2025

Digital Marketing & SEO Certification

Specialised track at JDC Free IT City — advanced search engine optimization and organic growth strategy.Growth

05 — Contact

Let's build something secure.

Security audit, AI pipeline, trading bot or a full product — I reply to everything within a day.

Whether you need to lock down an infrastructure, wire an intelligent AI agent, or just bounce technical ideas around — my inbox is always open.

umairghaffar759@gmail.com
Location
Lahore, Pakistan
Working
Remote-friendly · Worldwide
Response
Under 24 hours
Approach
Security-first
Open to opportunities
Local time
LHR --:--
Timezone
PKT · UTC+5

Delivered straight to umairghaffar759@gmail.com via a hardened FormSubmit AJAX relay. I reply within 24 hours.

umairghaffar759@gmail.com
✦Done