Data Scientist (Data & AI) - WFH

About the Company

Emory Healthcare supports patient care and healthcare operations through clinical expertise, technology, and data-driven decision-making. The organization also invests in employee development through professional mentorship, leadership programs, and a range of health and wellness benefits.

About the Role

Emory Healthcare is seeking a Data Scientist to transform clinical and operational data into reliable measures, predictive models, and AI-powered tools that support physicians, executives, and operational leaders. Working closely with clinical, finance, and operations teams, you will analyze questions related to patient length of stay, physician productivity, disease cohorts, readmission risk, and healthcare performance.

The role uses the Emory Unified Data Platform, including Epic Clarity and Caboodle data in Microsoft Fabric, alongside SQL, Python/PySpark, R, Power BI, and AI technologies. A strong focus on data validation is essential, ensuring that analytical results align with trusted reporting or that discrepancies are clearly explained.

Key Responsibilities

• Build and validate datasets and performance measures using Epic Clarity and Caboodle data.
• Define clinical cohorts and business metrics in collaboration with clinical and business stakeholders.
• Write SQL and PySpark code and validate results against existing reports before they are used for decision-making.
• Develop and deploy statistical and machine learning models for clinical and operational needs, including length of stay compared with GMLOS, readmission and transfer risk, patient throughput, and productivity.
• Build scheduled production pipelines in Microsoft Fabric and Azure, with appropriate monitoring and documentation.
• Apply natural language processing (NLP) and large language models (LLMs) to pathology reports, radiology reports, and clinical notes to extract structured information.
• Develop test datasets and evaluation processes for clinical text models, incorporating human review before results are relied upon.
• Create Power BI semantic models and reports that help users understand clinical and operational performance.
• Develop conversational data agents that answer leadership questions in plain language while applying appropriate row-level security.
• Protect patient information through minimum-necessary access, de-identification where possible, and compliance with HIPAA requirements.
• Respect the distinction between research and quality-improvement activities, including applicable IRB requirements.
• Communicate analytical results, metric definitions, limitations, and assumptions clearly to physicians and executives.
• Maintain concise, accessible documentation of analytical methods and findings.
• Promote reusable code, peer review, version-controlled definitions, and consistent analytical practices.
• Mentor analysts and support service-line operational leaders in using data to improve decisions and workflows.

Required Qualifications

• Bachelor’s degree; a master’s degree is preferred.
• At least five years of experience in healthcare or data analytics.
• At least two years of experience working for a healthcare system or payer organization.
• At least two years of experience working with relational databases.
• Proficiency in SQL and at least one additional programming language, such as Python, R, or SAS.
• Experience with statistical modeling, machine learning, and artificial intelligence.
• Excellent presentation and communication skills, including the ability to explain analytical findings to clinical and executive audiences.
• Experience in strategic planning and workflow optimization.
• Ability to mentor and coach analytics team members and service-line operational leaders.
• Strong attention to data accuracy, validation, and consistency with established reporting.

Preferred Qualifications

• Experience with Epic data platforms, including Clarity, Caboodle, or Cogito, and familiarity with clinical and revenue-cycle concepts such as encounters, DRG, GMLOS, wRVU, orders, and results.
• Experience with Microsoft Fabric and Azure, including lakehouses, Spark notebooks, Power BI semantic models, DAX, or Azure AI Foundry.
• Experience applying NLP or LLMs to pathology reports, radiology reports, or clinical notes, including model evaluation and human-in-the-loop review.
• Knowledge of protected health information (PHI) handling and the distinction between quality improvement and…