Data Scientist 2
About The Position
Optum Insight’s COE Advanced Analytics team is seeking a hands-on Data Scientist 2 to design, develop, deploy, and support enterprise-grade analytics, machine learning, Generative AI, and Agentic AI solutions for complex healthcare and operational use cases. The role combines data science, AI engineering, statistics, Python, SQL, machine learning, and cloud engineering, taking solutions from business problem definition and experimentation through production deployment, monitoring, and continuous improvement. The Data Scientist works with healthcare domain experts, product owners, data engineers, application engineers, platform teams, architects, and business stakeholders. The portfolio includes predictive and prescriptive analytics, patient-flow and operational forecasting, healthcare performance improvement, MLOps, model monitoring, natural-language analytics, Retrieval-Augmented Generation, multi-agent applications, and Responsible AI. UnitedHealth Group describes its mission as helping people live healthier lives and making the health system work better for everyone. It is committed to an inclusive workplace and is an equal employment opportunity and drug-free workplace; candidates must pass a drug test before beginning employment.
Requirements
- At least 2 years of professional experience delivering production-grade data science, machine learning, time-series forecasting, or AI engineering solutions.
- At least 2 years of hands-on Python experience for data analysis, model development, application integration, and production engineering.
- At least 2 years of SQL experience, including complex joins, aggregations, window functions, data validation, query optimization, and analytical dataset creation.
- At least 2 years of experience with machine learning libraries such as scikit-learn, XGBoost, PyTorch, or TensorFlow, and tools including Seaborn, Scikit-learn, and MLflow.
- At least 2 years of experience with distributed computing and cloud platforms, such as Apache Spark/PySpark, Azure, or Azure Databricks.
- At least 1 year of experience with APIs, application integration, Git, automated testing, CI/CD pipelines, and production support practices.
- Ability to use statistical and analytical techniques, including probability, hypothesis testing, regression, model evaluation, and experimental methods, to develop and evaluate robust solutions.
- Experience with model-performance monitoring, data-quality checks, data drift, model drift, logging, and troubleshooting.
- Ability to explain individual technical contributions, architectural decisions, implementation challenges, and measurable outcomes.
- Demonstrated experience operationalizing and deploying models to production beyond proof-of-concept stages.
- Candidates must pass a drug test before beginning employment.
Nice To Haves
- Experience delivering healthcare analytics or AI solutions in clinical operations, population health, patient flow, utilization, or financial performance domains.
- Deep experience with Azure Databricks capabilities such as Unity Catalog, Delta Lake, Workflows, Model Serving, MLflow, Databricks Apps, or Asset Bundles.
- Hands-on experience developing LLM, RAG, semantic search, vector retrieval, or Agentic AI applications using tools such as LangChain, LangGraph, or Semantic Kernel.
- Experience building REST APIs with FastAPI or Flask and containerized deployments using Docker or Kubernetes.
- Familiarity with Responsible AI practices, model interpretability, bias assessment, AI safety, and privacy-preserving analytics in regulated healthcare environments.
Responsibilities
- Translate complex healthcare, clinical, operational, and financial problems into analytical questions, hypotheses, metrics, and solution designs.
- Design, develop, evaluate, and optimize machine learning models for classification, regression, time-series forecasting, recommendation, and predictive or prescriptive use cases.
- Design and implement enterprise-grade Generative AI applications using large language models, embeddings, RAG, vector retrieval, prompt engineering, and multi-agent workflows.
- Develop Agentic AI and multi-agent workflows involving routing, planning, tool invocation, memory, reflection, and supervisor-worker orchestration.
- Implement guardrails, validation, traceability, access controls, and human-review mechanisms for enterprise AI solutions.
- Own the end-to-end lifecycle of machine learning and AI solutions, including experimentation, testing, deployment, MLOps, monitoring, drift detection, and continuous improvement.
- Build and maintain scalable data-processing workflows, feature tables, Delta tables, and model pipelines using Python, SQL, Spark, and Databricks.
- Apply software-engineering practices such as modular code design, Git-based version control, automated testing, code reviews, logging, and technical documentation.
- Apply statistical methods, hypothesis testing, and experimental design to structured and unstructured healthcare data, distinguishing correlation from causation.
- Ensure compliance with Responsible AI practices, including fairness evaluation, explainability, hallucination assessment, output validation, and secure handling of sensitive healthcare data.
- Partner with cross-functional technical and business stakeholders to break complex work into user stories, technical milestones, architectural designs, and actionable analytical findings.
Benefits
- Comprehensive benefits package, subject to eligibility requirements.
- Incentive and recognition programs, subject to eligibility requirements.
- Equity stock purchase, subject to eligibility requirements.
- 401(k) contribution, subject to eligibility requirements.
Stand Out From the Crowd
Upload your resume and get instant feedback on how well it matches this job.
What This Job Offers
Job Type
Full-time
Career Level
Mid Level
Education Level
No Education Listed
