Mahad A.

Key Skills

Software

Internal/Proprietary Tooling

Don't disclose

Top Subject Matter

3D LiDAR point-cloud perception and georeferenced mapping

Computer vision model training and evaluation (microscopy/optical image tasks)

AI system integration and computer-vision model training/validation

Top Data Types

3D Sensor

Image

Video

Top Task Types

Mapping

Classification

Object Detection

Fine-tuning

Computer Programming/Coding

Evaluation/Rating

Segmentation

Bounding Box

Text Generation

Freelancer Overview

Final Year Project: 3D LIDAR Scanning System (Python point-cloud pipeline for LAS and georeferenced 3D maps). Brings 3+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include laspy, ArcGIS Pro, and NumPy. Education includes Master of Science, NUST (SEECS) (2025) and Bachelor of Science, Information Technology University (ITU) (2025). AI-training focus includes data types such as 3D Sensor, Computer Code, and Programming and labeling workflows including Mapping, Evaluation, and Rating.

Labeling Experience

Python Engineer - Shahid Engineering Solutions

ImageClassification

As a Python Engineer, you deliver Python-based solutions across multiple concurrent client engagements focused on machine learning, deep learning, computer vision, and AI systems integration. You develop and validate ML/DL models for classification, detection, and image-processing tasks using PyTorch and numerical libraries while prioritizing correctness and reproducibility. You also bridge high-level Python workflows with performance-critical low-level code to meet hardware constraints. • Build and validate ML/DL models using PyTorch, NumPy, Pandas, and SciPy for end-to-end vision and AI workflows • Implement performance-critical routines in C and Assembly for optimized execution under hardware constraints • Support AI system lifecycle activities including data preprocessing, model evaluation, and deployment integration • Translate ambiguous client requirements into well-defined, testable engineering problems with reproducible documentation

2025 - Present

Python Engineer

Delivered machine-learning, deep-learning, and computer-vision solutions that required integrating end-to-end data preprocessing, model evaluation, and deployment for client systems. Developed and validated ML/DL models for classification, detection, and image-processing tasks with emphasis on correctness, reproducibility, and numerically reliable results. Bridged low-level performance constraints with higher-level Python workflows to support dependable model training and deployment. • Developed ML/DL pipelines for classification, detection, and image-processing tasks using PyTorch and numerical Python libraries. • Conducted model evaluation and validation to ensure reproducible, numerically sound outcomes. • Integrated AI capabilities into production workflows: preprocessing, evaluation, and deployment. • Translated ambiguous requirements into testable engineering problems with clear documentation to support dataset/model readiness.

2025 - Present

Final Year Project: 3D LIDAR Scanning System (Python point-cloud pipeline for LAS and georeferenced 3D maps)

3D SensorMapping

Developed and validated an end-to-end 3D point-cloud processing pipeline as part of a drone-mounted 3D LiDAR system. Converted raw range data into LAS point clouds and produced georeferenced 3D maps for downstream perception workflows. Ensured computational correctness and reproducibility using scientific Python tooling during data preparation and map generation. • Built a Python pipeline to convert raw LiDAR data into structured LAS point clouds (PCL, laspy). • Generated georeferenced 3D maps and integrated them into ArcGIS Pro outputs. • Supported 3D perception by preparing consistently formatted point-cloud datasets for analysis. • Verified results using numerical computing tools aligned with correctness-focused model/processing needs.

2024 - 2025

Technical Projects: AI-Based Optical Aberration Correction; Image Denoising & Super-Resolution

Implemented and delivered ML/DL computer-vision projects where training and validation required preparing data and evaluating model outputs. Trained deep learning architectures such as ResNet and U-Net for image correction, denoising, and super-resolution tasks. Focused on producing numerically sound, verifiable results through repeatable training and evaluation workflows. • Trained ResNet-18 for optical aberration correction and stitched microscopy images (PyTorch). • Built and trained a U-Net model for image denoising followed by an advanced upscaler for restoration. • Performed correctness-driven validation by verifying computational outputs during experimentation. • Prepared model-ready datasets through code-based preprocessing for computer-vision tasks.

2022 - 2025

Education

Work History