Data Scientist

About the role

The UHC Payment Integrity AI/ML Engineering Team at Optum is dedicated to preventing Fraud, Waste, and Abuse (FWA) in the healthcare claims ecosystem through advanced Machine Learning and Generative AI solutions.

Responsibilities

• Design, train, fine-tune, and deploy Large Language Models (LLMs) and Generative AI components for claims automation, anomaly detection, and investigative workflows
• Build and operationalize ML pipelines using Python, PySpark, and cloud-native architectures (Azure/AWS/GCP)
• Develop traditional machine learning models (classification, anomaly detection, NLP pipelines) for high-volume healthcare datasets
• Implement RAG (Retrieval-Augmented Generation) systems, embedding models, and vector database integrations
• Develop automated data processing, feature engineering, and model training pipelines using Spark, MLflow, Databricks, and big-data ecosystems
• Partner with product, engineering, and clinical domain teams to translate complex business challenges into scalable ML and GenAI solutions
• Optimize and monitor ML models in production, ensuring accuracy, latency, compliance, and responsible-AI best practices
• Present AI/ML solution designs, model insights, and GenAI architecture recommendations to technical and non-technical stakeholders
• Design, develop, and deploy AI-powered solutions to address complex business challenges with emphasis on responsible use of AI

Requirements

• Bachelor's degree in CS or IT related field
• 5+ years of hands-on experience in AI/ML engineering, deep learning, or applied machine learning
• 3+ years of experience in Python, PySpark, ML frameworks (TensorFlow, PyTorch), and distributed training
• 3+ years of experience with big-data systems (like Hadoop, Spark, Hive) and cloud platforms (like Azure, AWS, GCP)
• 2+ years of experience with LLMs, including: - Finetuning (LoRA, QLoRA, PEFT, SFT, or RLHF) - Prompt engineering & system design - RAG pipelines & vector search

Preferred Qualifications

• Prior experience with US healthcare datasets (claims, clinical, EMR/EHR, provider networks, payer ops)
• Experience deploying ML/LLM workloads using Databricks, MLflow, Kubernetes, or serverless inference
• Familiarity with modern GenAI tooling (LangChain, LlamaIndex, HuggingFace, OpenAI/Anthropic/Azure‑OpenAI APIs)
• Knowledge of deep learning architectures (Transformers, sequence models, contrastive learning)
• Experience optimizing model inference using quantization, distillation, or distributed GPU compute
• Demonstrated success in AI product delivery, cross-functional collaboration, and influencing technical strategy
• Strong grounding in ML fundamentals (feature engineering, model evaluation, A/B testing, MLOps best practices)

Benefits

Our comprehensive benefits package includes:

• Life resources and support
• Emotional education and tools
• Physical health plans and coverage
• Financial benefits for today and planning for the future, including retirement