Ali Haidar Email & Phone Number

Beirut, Beirut Governorate, Lebanon

• Developed and implemented a real-time camera calibration system with the Spinnaker FLIR API, enabling precise 3D reconstruction. This system facilitated accurate alignment of multiple camera views, significantly improving the quality of 3D models in real-time production.• Optimized a C++ multi-threaded pipeline for detecting and tracking foreign object debris (FOD) on airport runways using the YOLO framework and Kalman filters.• Automated the deployment process for AI models by developing bash scripts in Linux to build inference engines, ensuring seamless integration of AI models into production environments. Through this experience, I gained expertise in optimizing AI models using TensorRT for deployment on NVIDIA Jetson systems.• Gained extensive expertise in image preprocessing and utilized OpenCV to refine images for analysis, ensuring high-quality input for downstream AI models, thereby improving overall system accuracy and performance.

Beirut Governorate, Lebanon

• Developed a CSP entitled “Electronic Waste Mismanagement in Beirut Southern Suburbs” under the supervision of AUB CCECS, in partnership with Ecoserv NGO, and in collaboration with five fellow USAID scholars to promote sustainable development while utilizing our technical expertise.• Delivered awareness workshops about e-waste environmental and health hazards and performed field inspections to collect data for surveying and assessment.• Developed a prototype of an AI-powered recycling system for e-waste classification, utilizing the YOLO framework for object detection and integrating sensors and actuators with Arduino and Raspberry Pi.

Beirut, Beirut Governorate, Lebanon

• Developed electrical designs for MV/LV power supply and distribution networks, ensuring compliance with global standards and project specifications.• Designed earthing systems for enhanced safety and system reliability.• Collaborated with multidisciplinary engineering teams to optimize designs and align with client requirements.• Ensured project deliverables met both technical and regulatory standards while maintaining efficiency and cost-effectiveness.

Beirut Governorate, Lebanon

• Conducted measurements on vector network analyzer to record biosensors’ response to change in nicotine concentrations.• Measured dielectric variations with respect to changes in nicotine concentration over broad frequency spectrum using a dielectric probe, providing novel data.• Optimized input data by performing featuring engineering and utilizing feature forward selection wrapper method.• Developed a Gaussian Process Regression (GPR) model to predict nicotine concentrations from radio-frequency signals, particularly scattering parameter data, achieving 95% accuracy on a dataset of 100 real saliva samples.

Rouen, Normandy, France

• Performed image segmentation tasks on multi-institutional pre-operative MRI scans from the Brain Tumor Segmentation (BraTS 2020) Challenge dataset to segment heterogeneous brain tumors, particularly gliomas.• Acquired a comprehensive understanding of the BraTS20 dataset including its multimodal format (T1, T1ce, T2, FLAIR) and its segmentation labels of the brain tumor.• Explored and tested diverse preprocessing techniques applicable to 3-dimensional images, such as Numpy array conversion, scaling, cropping, transforming, data augmentation, and data splitting techniques. • Investigated the composition and building blocks of the U-net architecture, prominently used for segmentation tasks.• Researched and experimented with various loss functions employed for segmentation tasks, such as dice loss, binary cross-entropy, focal loss, and batch dice loss.• Implemented region-based training by modifying the segmented mask to encompass overlapping regions (namely whole tumor, tumor core, and enhancing tumor) instead of the default segmented labels. • Computed metrics on the overlapping regions instead of the default segmentation labels.Regarding software tools and implementation:• Became proficient in TensorFlow and PyTorch frameworks and their respective libraries and packages.• Managed remote clusters that operate using Slurm workload manager, utilized virtual environments, wrote batch scripts, specified GPU resources, and successfully submitted training jobs.

• Represented USAID scholars within the LEAD committee, successfully advocating for scholar-led initiatives and improvements.• Organized events and workshops that increased student participation by 80%, fostering stronger community bonds and enhancing professional development opportunities.

Bachelor Of Engineering - Be, Electrical And Computer Engineering, High Distinction Cgpa: 3.96/4.00

- Graduated with High Distinction, constantly being placed on the Dean’s Honor List. - Received the Dean’s Award for Creative Achievement.

Track, Control And Robotics, 4/4

Completed core modules including Control Systems, System Design and Analysis, System Identification, Instrumentation, and Machine.

High School Diploma, General Sciences, 18.11/20.00