Applied Data Scientist

Upperline Health•Nashville, TN

About The Position

Upperline Health is a provider group focused on lower-extremity, wound, and vascular care. It delivers coordinated, value-based specialty care through teams that include physicians, advanced practice providers, care navigators, pharmacists, dieticians, and social workers. The Applied Data Scientist, Model Pipelines will build, validate, deploy, and maintain predictive healthcare models. The role creates feature and model pipelines that transform claims and EHR data into predictions for applications such as risk stratification, utilization, expense and revenue projection, and outreach prioritization. The data scientist owns models from problem framing through monitoring, including reviewing historical prediction performance.

Requirements

Degree in Data Science or a related quantitative field, such as Statistics, Computer Science, Applied Mathematics, or Biostatistics.
2–5+ years of experience as a Data Scientist, Machine Learning Engineer, Data Engineer, or in a related role.
At least one model in production, with a focus on building and maintaining production-quality data pipelines and machine-learning infrastructure.
Proficiency in SQL/Databricks and Python.
Experience with retrospective analysis and backtesting, and evaluating model performance, calibration, bias, and drift.
Experience building scalable data pipelines, working with large datasets, and transforming raw data into reliable features for machine-learning workflows.
Strong working knowledge of machine-learning workflows such as scikit-learn, pandas, and NumPy, and Git; sound judgment about when a simpler model is preferable.
Strong problem-solving, communication, and critical-thinking skills, with attention to detail.
Comfort working in a fast-paced environment and flexibly managing multiple projects or analyses at once.

Nice To Haves

Master's degree in Data Science.
Background in healthcare analytics and experience with healthcare claims data, such as Medicare Claims, ACO REACH, or risk adjustment.
Experience with cloud platforms such as AWS or Azure, Databricks orchestration, and EMRs such as athenahealth.
Personal or work projects involving building and deploying machine-learning models to answer specific business questions.

Responsibilities

Design and implement SQL and Python (Databricks) data pipelines that synthesize large medical claims and EHR records into usable features on governed data.
Build, refine, and validate machine-learning models in Python, considering performance, calibration, bias, and drift in collaboration with analysts and engineers.
Conduct retrospective lookback analyses by backtesting predictions against historical claims and outcomes, and measure whether outreach or interventions changed results.
Deploy models using shared engineering tooling and manage model artifacts.
Monitor models and retrain them as needed.
Ensure data quality, reproducibility, and performance in pipelines.
Contribute to data-governance and model-deployment best practices.
Explain results to nontechnical leaders and connect them to actions.