CV
Contact Information
| Name | Osama Ahmad |
| Professional Title | PhD Student |
| oahmad@umass.edu |
Professional Summary
PhD student in Electrical and Computer Engineering at UMass Amherst. My research interests include LLM systems, healthcare AI, graph neural networks, computer vision, embedded systems, and intelligent transportation systems.
Experience
-
2026 - present Amherst, MA, USA
Graduate Research Assistant
UMass Amherst
- Exploring and developing LLMs for healthcare applications.
-
2025 - 2026 Australia
Data Engineering Research Assistant
Maincode
- Designed scalable data crawling, filtering, and deduplication pipelines to curate high-quality datasets for large language model training.
-
2025 - 2026 Lahore, Pakistan
Adjunct Faculty — AI on Edge Devices
Lahore University of Management Sciences
- Delivered graduate-level lectures and hands-on labs on TinyML, model compression, and on-device inference.
-
2025 - 2025 Lahore, Pakistan
Graduate Teaching Assistant — Machine Learning
Lahore University of Management Sciences
- Designed practice problems and coding assignments in Python.
- Graded assignments, quizzes, and projects with detailed feedback.
-
2021 - 2026 Lahore, Pakistan
Research Assistant
Lahore University of Management Sciences
Worked at the Machine Vision and Artificial Intelligence Lab and Center for Urban Informatics, Technology, and Policy.
- Designed IoT-based smart vending machine systems using embedded sensors.
- Developed surface defect detection algorithms using camera-based image processing.
- Built ANPR-based speed enforcement pipelines for intelligent transportation systems.
- Worked on average speed monitoring and smart parking systems.
-
2020 - 2020 Lahore, Pakistan
Research and Development Engineer
Midas International
- Designed embedded systems for condition monitoring devices.
- Developed and calibrated sensors for HVAC systems.
-
2019 - 2019 Lahore, Pakistan
Project Engineer
Dimen Draw
- Prototyped a safety airbag jacket for motorbikes.
- Designed a sensor-selection tool based on research methodologies.
Education
-
2026 - present Amherst, MA, USA
PhD Student
University of Massachusetts Amherst
Electrical and Computer Engineering
-
2022 - 2024 Lahore, Pakistan
Master of Science
Lahore University of Management Sciences
Electrical Engineering
- Dean’s Honour
- Thesis: Dynamic Decoupling of Spatio-temporal Data in Graph Networks for Traffic Forecasting and Beyond
-
2015 - 2019 Lahore, Pakistan
Bachelor of Science
University of Engineering and Technology Lahore
Mechatronics and Control Engineering
- Undergraduate Thesis: 3D Human Limb Profile
Publications
-
Review Robust Spatiotemporal Forecasting Using Adaptive Deep-Unfolded Variational Mode Decomposition
IEEE Signal Processing Letters [Under Review]
-
Review Variational Mode-Driven Graph Convolutional Network for Spatiotemporal Traffic Forecasting
ACM Transactions on Intelligent Systems and Technology [Under Review]
-
2025 Robust and Noise-resilient Long-Term Prediction of Spatiotemporal Data Using Variational Mode Graph Neural Networks with 3D Attention
IJCNN
Skills
Programming: Python, C++, JavaScript, PHP
AI / ML: PyTorch, OpenCV, TensorFlow.js, Flask, MATLAB
Embedded Systems: Arduino, PIC, ESP32, Raspberry Pi, Mini6410, myRio, MPLAB
Tools & Technologies: Gazebo, ROS, Node.js, SolidWorks, PCB designing, IoT
Projects
-
IoT-Based Smart Vending Machine
Designed and developed an IoT-enabled smart vending machine using embedded systems and beam-break IR sensors for falling-object detection.
- Developed embedded firmware for sensor interfacing and real-time control.
- Designed IR beam-break sensing mechanisms.
- Implemented sensor-data filtering and noise reduction.
- Integrated IoT communication modules for monitoring.
-
Surface Defect Detection System
Developed a computer-vision-based inspection system for detecting surface defects on packaging materials.
- Designed image acquisition and preprocessing pipelines.
- Implemented defect detection algorithms using OpenCV and Python.
- Improved throughput using multiprocessing and multithreading.
- Reduced false detections under varying illumination conditions.
-
Intelligent Speed Enforcement System
Designed a deep-learning-based ANPR system for vehicle speed enforcement in dynamic traffic environments.
- Developed an automatic number plate recognition pipeline.
- Integrated radar-based speed estimation.
- Optimized OCR and inference latency for real-time deployment.
- Improved robustness under motion blur and lighting variation.