Email copied to clipboard: shakeelrana6240@gmail.com
Available for Hire · 2026 CS Grad

SHAKEEL
UR REHMAN

> Computer Vision Specialist

Building production-ready Computer Vision pipelines, autonomous military-grade telemetry defense systems, and high-throughput Applied ML models. Final-year Computer Science student at UET Lahore.

15+
CV & ML Projects
<2.5s
IoV Dispatch Latency
99.2%
Tracking Precision
YOLO11 + SAHI
LangGraph Multi-Agent
Shakeel Ur Rehman - AI Engineer
Shakeel Ur Rehman
AI / ML Engineer · UET Lahore
YOLO11 DeepSORT PyTorch SHAP IoV

Real-Time Perception & Telemetry HUD

Test drive simulated edge telemetry pipelines directly in your browser. Switch between my core project domains: Anti-Drone Defense, Smart IoV emergency response, and tabular Churn explainability.

model_runtime_hud.v11 — ShakeelRana624
FPS: 60.0
LATENCY: 14.2 ms
STATUS: 1 Target (LOCKED)
Telemetry Data Stream ● SYNCED

Bridging Theory & Production in
Applied AI & Computer Vision

I am a final-year Computer Science student at University of Engineering and Technology (UET) Lahore (graduating 2026), dedicated to designing and deploying resilient artificial intelligence systems.

My work focuses heavily on real-time video intelligence (combining YOLO11/v8 with SAHI for small-object aerial detection and DeepSORT for multi-target tracking), explainable predictive modeling (XGBoost + SHAP feature importance), and agentic workflows (LangGraph multi-agent systems with sub-second dispatch coordination).

I believe that building real AI means architecting the complete lifecycle: clean exploratory data analysis, robust training pipelines, containerized microservices via FastAPI & Docker, and intuitive interfaces for end-users.

🎓
UET Lahore (2022–2026)
BS Computer Science · Final Year
🛰️
Autonomous UAV & IoV
Sub-2.5s Emergency Dispatch Pipeline
⚡
Production & MLOps
FastAPI, Docker, Streamlit & Full-Stack
💼
Open for Opportunities
Remote Contracts, Freelance & Full-Time
01
Real-Time Perception & Aerial Tracking
Leveraging Sliced Aided Hyper Inference (SAHI) and DeepSORT over state-of-the-art YOLO detectors to track fast-moving objects, drones, and humans in complex multi-scale environments.
02
Explainable Applied Machine Learning
Engineering high-accuracy tabular and classification pipelines using Optuna-optimized gradient boosters (XGBoost, LightGBM, CatBoost) grounded in SHAP explainability.
03
Agentic AI & Connected Systems (IoV)
Architecting multi-agent coordination pipelines using LangGraph and Internet-of-Vehicles (IoV) protocols for automated emergency incident response and cognitive agent memory.

Technologies & Frameworks

Tools, libraries, and architectures I leverage daily to construct high-performance machine learning models and computer vision pipelines.

👁️

Computer Vision & Perception

Real-time object detection, small-target slicing, spatial tracking, and human pose estimation.

YOLO11 YOLOv8 SAHI DeepSORT OpenCV YOLOv8-Pose MediaPipe CLIP
🤖

Applied ML & Deep Learning

Deep neural networks, tabular predictive models, automated hyperparameter tuning, and model explainability.

PyTorch TensorFlow Scikit-Learn XGBoost LightGBM CatBoost SHAP Explainability Optuna
🧠

Agentic AI & Generative Models

Autonomous multi-agent swarms, cognitive architectures, RAG pipelines, and conversational interfaces.

LangGraph LangChain Claude API Gemini API RAG Vector DBs LAMB-Core Multi-Agent AI
⚙️

MLOps & Full-Stack Deployment

Containerization, REST APIs, asynchronous messaging, and end-to-end cloud production delivery.

FastAPI Docker Python Streamlit Firebase Next.js Git / GitHub Weights & Biases

Selected Engineering Projects

A showcase of deployed applications, computer vision systems, and applied machine learning research.

🧠 AI / ML · DEFENSE & COMPUTER VISION
Military-Grade

Autonomous Anti-Drone Telemetry Defense System

An autonomous, AI-driven military-grade Anti-Drone Detection & Early Warning System featuring real-time telemetry tracking. Automatically calculates target GPS coordinates (Lat/Lon), true compass bearing, speed, altitude, and distance estimation utilizing YOLO11, SAHI (Sliced Aided Hyper Inference) for small airborne targets, and DeepSORT tracking.

YOLO11 SAHI DeepSORT OpenCV GPS Telemetry PyTorch
🧠 AI / ML · APPLIED ML & SHAP
Full-Stack ML

Telco Customer Churn Prediction & Explainability Engine

End-to-end Machine Learning pipeline for telecommunications customer churn prediction. Features rigorous Exploratory Data Analysis, Bayesian Hyperparameter Tuning with Optuna, SHAP (SHapley Additive exPlanations) for model interpretability, and a production FastAPI web application containerized with Docker.

XGBoost Scikit-Learn SHAP Optuna FastAPI Docker
🧠 AI / ML · POSE ESTIMATION & CV

RealTime Pose Estimation & Activity Detection Engine

Modular deep learning-based human skeleton pose estimation and activity detection engine built with YOLOv8-Pose for real-time multi-person postural analysis, joint angle calculation, and ergonomic tracking.

YOLOv8-Pose PyTorch OpenCV Skeleton Keypoints
🧠 AI / ML · GENERATIVE COMPUTER VISION

AI Virtual Try-On System (H-VITON)

Deep learning virtual garment synthesis leveraging high-resolution H-VITON neural networks. Warps clothing realistically across arbitrary 2D body poses while preserving textures, fabric folds, and patterns.

PyTorch H-VITON Neural Warping GANs OpenCV
🧠 AI / ML · CYBERSECURITY & NLP

Adversarial Phishing Attack Detection Classifier

High-precision machine learning classifier for detecting malicious URLs and phishing email payloads using lexical NLP analysis, adversarial feature extraction, and gradient-boosted decision trees.

XGBoost NLP Scikit-Learn Feature Extraction
🧠 AI / ML · COGNITIVE AGENTIC AI

LAMB-Core: Cognitive Memory Operating System

Enterprise-grade cognitive memory architecture for autonomous AI agents, featuring human-inspired attention models, working memory consolidation, associative vector retrieval, and lifelong adaptive learning.

Agentic AI Cognitive Memory Vector DB Python
🧠 AI / ML · AGENTIC LLM & SAAS

AI Resume Analyzer Pro (Agentic Full-Stack SaaS)

Production SaaS application featuring agentic AI resume parsing, ATS scoring algorithms, Claude API intelligence, Stripe checkout subscription integration, and containerized Docker architecture.

Claude API Agentic AI Next.js Stripe Docker Python
🧠 AI / ML · AGENTIC CONVERSATIONAL AI

Smart Restaurant Conversational AI Assistant

Autonomous agentic ordering assistant powered by LangChain and Google Gemini with live menu inventory awareness, context-preserving conversation memory, and automated order confirmation.

LangChain Agentic AI Gemini API FastAPI Python
🌐 WEBSITES · 3D MOTION DESIGN

Limon's Agency: Interactive 3D Product Showcase

Premium mobile app development agency website featuring interactive 3D WebGL product visuals, high-performance smooth animations, bespoke motion design system, and responsive layout.

HTML5 / CSS3 Three.js Motion Design WebGL
🌐 WEBSITES · E-COMMERCE & 3D PARALLAX

Ooga Booga: Immersive Vintage-Cartoon Experience & Store

Custom immersive vintage-cartoon theme featuring multi-layered parallax artwork, interactive 3D product experiences, kinetic motion design, and responsive layout.

Liquid CSS3 Parallax 3D Interactions JavaScript

Have an AI Challenge?
Let's Engineer The Solution.

Available for freelance contracts, remote consulting, and full-time AI/ML & Computer Vision engineering roles. Guaranteed response within 24 hours.

✦ Initiate Collaboration