Optum is a global organization that delivers care, aided by technology to help millions of people live healthier lives. The work you do with our team will directly improve health outcomes by connecting people with the care, pharmacy benefits, data and resources they need to feel their best. Here, you will find a culture guided by inclusion, talented peers, comprehensive benefits and career development opportunities. Come make an impact on the communities we serve as you help us advance health optimization on a global scale. Join us to start Caring. Connecting. Growing together.
The Provider Enablement Methods and Data Science team is seeking a Senior Data Scientist to support various Optum Insights data science projects. This position is responsible for end-to-end model development for our clients in provider and payer segments. We are investing significantly in new capabilities to design, develop and deliver better solutions for these clients. We are seeking people with proven experience conceptualizing, defining and delivering solutions that enable data science teams to build and scale Machine Learning and Artificial Intelligence models that will power these solutions.
Primary Responsibilities:
- Build and put into production LLM solutions across major use cases (Q&A, generation, summarization, classification, conversational/assistant, and multimodal)
- Apply modern LLM development techniques including PEFT (LoRA/adapters), RAG, prompt design, tool/function calling (incl. MCP patterns where relevant), and structured output enforcement
- Establish robust LLM evaluation and quality programs using automated metrics plus structured human review, covering relevance, faithfulness, hallucinations, robustness, bias, readability, and offline/online evaluation strategy
- Implement Responsible AI guardrails including promptinjection defense, toxicity and safety filtering, privacy/PII controls, scope/refusal behavior, adversarial testing, and mitigation of automation bias in review processes
- Design and optimize retrieval and vector search workflows (chunking, indexing, reranking, grounding, and handling lowsignal/conflicting context)
- Orchestrate multistep AI workflows with reliable control logic (branching, retries, tool execution, state/graph orchestration) integrated with evaluation gates and guardrails
- Lead endtoend ML delivery from data preparation and feature engineering through modeling, evaluation, deployment, monitoring, and iterative improvement using reproducible practices
- Build scalable data foundations using advanced SQL and distributed processing, ensuring correctness, performance, and stability for large multisource healthcare datasets
- Develop advanced modeling solutions across statistics, ML, deep learning, and NLP (including NER, OCR to text pipelines, LSTM/sequence models) with solid Transformer proficiency and clear method tradeoff reasoning
- Own MLOps practices including model lifecycle management, CI/CD releases, monitoring and drift detection (data/concept), structured peer review/validation, data pipelines/orchestration, governance, architecture/design documentation, healthcare analytics application, agile execution, and solid Python engineering practices
- Comply with the terms and conditions of the employment contract, company policies and procedures, and any and all directives (such as, but not limited to, transfer and/or re-assignment to different work locations, change in teams and/or work shifts, policies in regards to flexibility of work benefits and/or work environment, alternative work arrangements, and other decisions that may arise due to the changing business environment). The Company may adopt, vary or rescind these policies and directives in its absolute discretion and without any limitation (implied or otherwise) on its ability to do so
Required Qualifications:
- A Bachelor's degree in Data Science, Statistics, Mathematics, Computer Science, Machine Learning, Economics, Engineering, or a related quantitative field, or equivalent practical experience
- 8+ years of hands-on experience designing, developing, and validating statistical, machine learning, and/or deep learning models in production or applied research settings
- 8+ years of experience working with largescale structured and unstructured datasets, preferably within healthcare, life sciences, financial services, or other regulated domains
- Experience designing and evaluating models using appropriate performance metrics, validation strategies, and error analysis techniques to ensure robustness and reliability
- Experience communicating complex technical concepts, modeling results, and analytical insights to both technical and nontechnical stakeholders through clear documentation and presentations
- Demonstrated experience applying statistical methods, machine learning algorithms, and computational techniques to extract insights and build predictive solutions from complex datasets
- Solid proficiency in core data science and engineering tools, including Python, SQL, and distributed processing frameworks such as Spark or equivalent technologies
- Demonstrated ability to work collaboratively within cross functional teams while also operating independently and taking technical ownership of complex data science initiatives
- Proven ability to translate ambiguous business or product questions into clearly defined analytical problems, end to end modeling approaches, and measurable success criteria
Preferred Qualifications:
- Experience with containerization practices and tools such as Docker, including packaging ML/LLM services for reproducible deployment and environment consistency
- Experience with orchestration platforms such as Kubernetes for deploying and managing scalable ML/LLM workloads in production environments
- Experience with cloud services and managed ML platforms for deployment and operations, including designing cloud native architectures that balance latency, reliability, and security requirements
- Experience with ML Ops tooling such as ML flow, Kubeflow, and/or TensorFlow Extended (TFX) for experiment tracking, pipeline automation, model registry workflows, and production governance
- Experience implementing robust evaluation and monitoring for GenAI systems beyond development environments, including post deployment observability, drift detection for prompts and responses, and operational alerting for quality regressions
- Experience building and maintaining cross platform data and model infrastructure, including cost/performance optimization decisions across compute and storage
- Experience with deploying ML models in Azure, AWS, and/or Google Cloud
