2026-7843 Data Scientist-Senior (Midshift)

The ideal Senior Data Scientist candidate for this role will have insurance industry experience with Natural Language Processing (NLP) and experience designing and developing agentic workflows and AI-driven systems, especially applied to data extraction problems. They will be skilled in transfer learning and possess an understanding of and capability in Deep Learning (DL). They can design and implement autonomous or semi-autonomous agent-based solutions, orchestrating multi-step reasoning, tool usage, and decision-making workflows.

They can perform work in each of those areas by leveraging a cloud environment like Databricks on Azure with Python in a notebook and IDE environment with version control leveraging git. They can navigate API-based generative AI models like those from OpenAI and similar providers to develop agentic, workflow-driven solutions. They can perform these tasks collaboratively, transparently, and seek to improve the skill of the team overall.

Responsibilities

• Deliver production-quality solutions for data extraction, classification, triaging, routing, search, and agentic workflow orchestration
• Design, build, and optimize agent-based systems that can plan, reason, and execute multi-step tasks using tools and APIs
• Programmatically explore data, derive insights that are statistically sound, and convey findings to colleagues of varying sophistication
• Develop and maintain workflow pipelines that integrate LLMs, external tools, and business logic for scalable automation
• Label and validate datasets of various sizes, using manual methods as well as programmatic and AI-driven approaches
• Evaluate the performance of models and end-to-end agentic systems using practical and statistical benchmarks
• Collaborate with other scientists, engineers, product owners, and business customers to develop solutions that meet the business problem
• Create, contribute to, improve, and convey technical and non-technical documentation of solutions

• Experience applying quantitative methods in a corporate environment
• Experience with Python from a functional programming paradigm, able to manage dependencies, virtual environments, and version control in git
• Experience with cloud computing platforms such as Azure
• Expertise in supervised and unsupervised learning along with experience in deep learning and transfer learning
• Experience designing and implementing agentic workflows, including chaining, tool usage, memory management, and decision logic
• Experience working with generative AI and foundation models (e.g., GPT-4o, Mistral), including prompting, orchestration, and evaluation of agent behaviors
• Experience developing solutions from inception through deployment

• Graduate degree in a quantitative field
• Experience with sequential algorithms (e.g., LSTM, RNN, GRU, etc.)
• Experience building or contributing to agent frameworks (e.g., LangChain, Semantic Kernel, AutoGen, or similar orchestration layers)
• Experience evaluating ethical implications of AI and considerations around controlling them