Shoaib Zafar zafar

About

I am Machine Learning / Data Engineer passionate about building scalable AI solutions across LLMs, computer vision, medical imaging, and MLOps. Experienced in developing production-ready ML pipelines, fine-tuning Transformer-based models with RLHF, and deploying AI systems using Docker, MLFlow, Airflow and Kubeflow CI/CD tools. At 10Pearls, I built an AQI prediction system with an ETL pipeline automated MLOps pipeline achieving 95% accuracy with Ensemble learning which selects and Daily retrains the model daily predicts every 3 hours. It reduces the risk of hazardous prediction by 50% and increase the efficiency and reliability by 60%. I developed a GPU-optimized YOLOv11 computer vision Object Detection model solving Fleet management. I collect 4995 images on highway traffic created a Custom Highway Dataset trained the Object Detection model on RTX 4050 GPU with CUDA and cuDNN with 0.85 accuracy 0.92 precision recall and improved the efficiency about 45%. NeuroVision.AI a Robust Innovative AI Research project on MRI Medical imaging using GenAI a finalist in the Ignite AI Wrapper Competition 2025 . an MRI brain tumor segmentation and 3D Neural reconstruction solution built using MONAI, U-Net, and GANs. It revolutionizes the way we diagnosie disease like brain tumor,liver tumor improving efficiency by 25% percent reducing the cost by 40%

Experience

• Data/AI Engineer

  Huzzle

  7 July 2026 – Present

Find online

Subscribe to access direct links and contact information.