Architecture, Software Engineering and AI/ML Engineering for production grade Machine Learning models to be deployed in vehicles. Development of model and data interfaces, Pipelines to train and evaluate models on the Azure cloud. Conduct ML experiments, generate metrics and track artifacts using Weights and Biases. Quantization and Release for Qualcomm-based embedded platform. Integration within larger group to develop an Active Learning Pipeline for the Big Data Loop using multiple models and to detect and reduce model uncertainty. Quality assurance using
automated tools and pipelines.
Keywords: Python 3, Pytorch, Pytorch lightning, torchvision, Qualcomm/QNN, Weights&Biases (wandb), albumentations, Semantic Segmentation, Object Detection, YoloX, torchmetrics , ONNX, OnnxRuntime, OpenCV, Tensorflow, scikit-learn, NumPy, pandas, pytest, Mlflow, Conda, Anaconda, Mamba/Micromamba, Nvidia, CUDA, Azure Cloud, Azure DevOps, Azure Pipelines, Docker, terraform, packer, git, LFS, pre-commit, jupyter