Data Scientist

Full-time Mathematical Science Occupations

<ul> <li>Rapidly prototype and benchmark multiple computer vision approaches for anomaly and defect detection, including supervised detection and classification, unsupervised and semi-supervised anomaly detection, and modern foundation model based techniques on representative inspection imagery</li> <li>Build fair and reusable evaluation frameworks so that comparisons across methods are repeatable and credible across different asset types and defect categories</li> <li>Identify top-performing methodologies and clearly document trade-offs including data requirements, performance, complexity, and inference cost</li> <li>Package the selected approach into a reusable template, including reference code, data preparation guidance, training and evaluation patterns, and supporting documentation so internal teams can extend it to additional use cases</li> <li>Collaborate closely with data scientists, domain experts, and stakeholders ...