Software Engineer - Machine Learning (SDV), POWER

Key Responsibilities

• Design, train, and deploy machine learning models tailored for resource-constrained edge hardware.
• Apply data science methodologies to perform Exploratory Data Analysis (EDA), statistical modeling, and visualization of complex time-series data from powertrain sensors to inform model architecture.
• Perform feature engineering to clean, process, and structure raw sensor datasets, ensuring they are optimized and ready to be fed into the models.
• Conduct hyperparameter tuning to continuously optimize machine learning models for higher accuracy and peak performance.
• Optimize model inference and runtime predictions to meet strict real-time execution constraints on target edge hardware.
• Develop real-time application logic using C and C++ while leveraging Python for model training and data preprocessing.
• Optimize ML models using techniques such as quantization (e.g., float32 to int8), model pruning, and memory optimization.
• Process, filter (e.g., using Kalman filters), and synchronize noisy data acquired from physical temperature and current sensors.
• Interface with and customize hardware platforms, including Raspberry Pi capabilities, Linux/Raspbian OS, GPIO, and SPI.
• Test, benchmark, and validate the performance of ML estimators against real hardware setups or high-fidelity powertrain simulators.
  
  

• Proven experience in feature engineering, dataset preparation, and data pipeline development.
• Solid background in applied data science, statistical analysis, anomaly detection, and data visualization for time-series sensor data.
• Strong knowledge of hyperparameter tuning methodologies to maximize model accuracy and efficiency.
• Deep understanding of model inference optimization and low-latency runtime prediction behavior.
• Strong programming skills in C, C++, and Python.
• Deep understanding of Raspberry Pi hardware, Linux/Raspbian OS customization, and embedded hardware interfaces (GPIO, SPI).
• Practical experience in model optimization, quantization, and pruning to meet strict hardware constraints.
• Solid background in signal processing and handling noisy data from physical sensors.
• Experience validating ML models against real hardware or advanced simulators.
• Experience in TensorFlow Lite for Microcontrollers & SciKit Learn is a plus
  
  

• 3+ years of relevant experience in machine learning, edge computing, or embedded software development.