Senior Data Scientist

Grade Level (for internal use): 10 Position Summary We are looking for a hands-on Data Scientist to join the Commodities At Sea (CAS) team. The role combines large-scale data processing, data wrangling and analytical problem-solving with the development of reliable, production-ready data products for maritime and commodity-flow analytics. The successful candidate will work extensively with Python and PySpark, use R where appropriate, and collaborate with product managers, industry analysts, data scientists and technology partners. The role also contributes to cloud and DevOps activities across AWS, Docker and Terraform when needed. A key part of this position is applying generative AI and intelligent workflow automation to improve how the team discovers, validates, transforms and operationalizes data. This includes building secure, observable and human-governed AI-assisted workflows, evaluating their quality, and integrating them responsibly into existing data pipelines and business processes.

Data Science, Engineering & Quality Design, build and maintain scalable data-processing and analytical workflows using Python, PySpark and SQL; use R where it provides a clear analytical advantage. Collect, clean, reconcile and transform large, complex, disconnected and semi-structured datasets into analysis-ready and product-ready assets. Develop reusable data components, features, models, validation checks and monitoring that improve the quality, reliability and explainability of CAS outputs. Perform exploratory analysis and root-cause investigation to identify anomalies, answer business questions and recommend data or process improvements. Build and optimize extraction, transformation and loading workflows using AWS data services and distributed-computing technologies. Create fit-for-purpose analytical tools, prototypes and visualizations that help product, research and data-science stakeholders make informed decisions. AI Workflow Automation Identify high-value opportunities to automate repetitive data and research workflows using generative AI, machine learning and rules-based orchestration. Design and implement AI-assisted workflows such as document and data extraction, semantic classification, summarization, entity matching, anomaly triage, metadata enrichment, code assistance and natural-language interfaces to curated data. Build workflow components using large language model APIs, prompt templates, structured outputs, retrieval-augmented generation (RAG), tool/function calling and agentic patterns where appropriate. Integrate AI components with data pipelines, APIs and business processes, with clear human review points for consequential or low-confidence outputs. Create evaluation datasets and automated tests for accuracy, relevance, grounded ness, robustness, latency and cost; monitor production behavior and continuously improve prompts, models and workflows. Apply responsible AI and security-by-design practices, including access control, data minimization, secrets management, auditability, traceability and protection against prompt injection or unintended data exposure. Document AI workflow design, assumptions, limitations and operating procedures, and enable colleagues to use approved AI capabilities effectively. Cloud, DevOps & Delivery Package workloads in Docker containers and contribute to reproducible development and deployment practices. Support AWS-based processing, storage, orchestration, monitoring and troubleshooting for analytical workloads. Contribute to infrastructure-as-code using Terraform and to CI/CD practices for automated testing, scanning and deployment. Work with stakeholders to define deliverables, manage dependencies and risks, and take ownership of quality and timely delivery. Support production incidents and continuous improvement through structured troubleshooting, observability and root-cause analysis. Business & Domain Contribution Develop a strong