About this role
Lead end-to-end delivery of predictive analytics solutions, including problem definition, data preparation, feature engineering, model development, validation, and deployment. Independently translate business objectives into analytic approaches, success metrics, and project plans for medium-sized data science initiatives. Partner with cross-functional stakeholders to gather requirements, prioritize use cases, and align analytics deliverables with business strategy. Apply machine learning, deep learning, statistical modeling, and optimization techniques to large structured and unstructured datasets to generate insights and predict outcomes. Identify and test hypotheses using appropriate statistical methods and evaluate model performance to ensure statistical rigor of findings. Develop and apply computer vision methods for insurance-domain use cases, including model training, evaluation, and monitoring. Produce clear visualizations and written/presented narratives that communicate findings and recommendations to non-technical and senior audiences. Document modeling methodology, assumptions, limitations, and results to support reproducibility and governance and implement monitoring to maintain model performance over time. Provide technical guidance through code review and best-practice recommendations to support team execution and consistency. Employer will accept a PhD degree in Computer Science, Statistics, or related field and two (2) years of experience in the job offered or in an Asst Dir, Data Science (STP)-related occupation. Alternatively, employer will accept a Master's degree in Computer Science, Statistics, or related field and four (4) years of experience in the job offered or in an Asst Dir, Data Science (STP)-related occupation. Conduct experiments and data-driven studies to support business decisions in an insurance setting, including hypothesis testing, interpreting results, and communicating findings. Deliver strategic decision support using insurance analytics and pricing platforms, including Emblem, Earnix, or equivalent tools. Implement data and model artifact versioning and reproducible ML workflows using version control tools, including DVC and MLflow. Build and maintain monitoring dashboards and production data pipelines using data warehouse and workflow orchestration tools, including Snowflake and Airflow. Apply geospatial fundamentals, including coordinate reference systems, map projections, and pixel-to-geo transforms. Train, evaluate, and improve computer vision models, including object detection and segmentation, and perform model performance diagnostics and error analysis. Optimize deep learning training performance through profiling, efficient data loading, and training optimizations. Develop, deploy, and operate data science solutions using public cloud platforms, including AWS and Azure. Containerize ML services and pipelines using Docker and apply CI/CD and deployment automation practices for production releases. Monitor and troubleshoot production scoring and data pipelines, including performance tracking, incident triage, and root-cause analysis. Develop and maintain labeling guidelines and label taxonomies for computer vision datasets, and coordinate with annotation resources to support dataset development. Experience may be gained during graduate program. Will accept any suitable combination of education, training, and/or experience. Multiple Positions Available.
