6+ years delivering production-grade systems across cybersecurity, precision agriculture, healthcare, and defense. From national-scale platforms to agentic AI workflows — I build things that ship and scale.
I'm a Senior Full Stack Developer and AI Engineer with over 6 years of experience delivering end-to-end platforms across government, healthcare, defense, and cybersecurity — from pixel-perfect frontends to production-grade agentic AI backends.
My work spans national-scale deployments like Agriverse (Pakistan's indigenous precision agriculture platform), hospital-deployed AI at Al-Shifa International Hospital (Roshan AI), and enterprise security intelligence with Cyber Nexus.
I specialize in building intelligent agentic systems using LangChain, LangGraph, and local LLMs — systems that don't just respond, but reason, plan, and act autonomously.
Production-grade threat intelligence platform with MITRE ATT&CK mapping, OSINT-driven LLM agents, live vulnerability feeds, and RAG-powered threat correlation.
Pakistan's indigenous national-scale platform performing satellite imagery inference for crop analysis, disease detection, and yield predictions on Kubernetes.
Clinical AI platform deployed at Al-Shifa International Hospital serving millions of patients. Integrates AI models with maternal health records for early GDM detection.
Company web platform driving traffic to Mobify's game business, featuring Polaris — a real-time AG Grid + Django REST dashboard for live business data monitoring.
LangGraph agentic pipeline using Qwen-VL for invoice extraction. Confidence-check nodes re-examine uncertain fields. Human-in-the-loop for low-confidence extractions.
Ingests image collections, generates semantic descriptions via LLaVA/Qwen-VL, stores in FAISS/ChromaDB, and answers natural language questions by reasoning over retrieved images.
Multi-agent system using Swarm + Ollama for autonomous CV parsing, candidate scoring, and structured hiring recommendations — entirely local, zero API cost.
Fully voice-operated LangGraph agent for monthly expense management. Runs entirely offline on local models — no data leaves the device.
RAG chatbot trained on a large corpus of game walkthroughs using ChromaDB and nomic-embed-text. Answers player questions with contextually retrieved sections via Ollama.
Scores uploaded CVs against criteria, matches with job openings, and returns structured improvement suggestions combining document intelligence and job market retrieval.
Django app that accepts a YouTube URL, extracts the transcript, and generates a publish-ready blog post via LLM — fully automated, no manual editing.
Private, fully offline agentic application for drafting professional emails via guided multi-turn interaction. Runs on local models — zero cloud dependency.
Production-extensible real-time chat using FastAPI WebSockets backend and browser-native WebSocket API. Clean chat-style UI, extensible starting point for production systems.
Custom POS for a plywood business with inventory tracking, expense calculation, billing, auto stock recalculation, and performance graphs — fully offline desktop solution.
Semantic segmentation on the Dynamic Earth satellite dataset evaluating all three spatial configurations (no shift, spatial shift, temporal shift) for deep learning model robustness analysis.
Deep learning-based autonomous vehicle steering control using computer vision and neural networks in a simulated environment — applied ML for autonomous systems.
Andrew Ng · Coursera
FreeCodeCamp
FreeCodeCamp
FreeCodeCamp
IBM · Coursera
IBM · Coursera
FreeCodeCamp
Actively looking for Senior Full Stack / AI Engineering roles in Riyadh, Saudi Arabia. If you're building something that needs production-grade full stack development, agentic AI systems, or cloud-native architecture — let's connect.