Data Scientist

Role Overview

Fractal is seeking a mission-driven Data Scientist to design, develop, and deploy AI and machine learning solutions that improve Healthcare Revenue Cycle Management (RCM) workflows. The role focuses on predictive models, generative AI, and intelligent automation to increase efficiency across claims processing, denials and appeals, clinical documentation, coding, and customer service.

Responsibilities

• Develop machine learning, deep learning, and generative AI models for RCM use cases such as:
  • Claim outcome prediction
  • Denials classification
  • Next best action recommendation
  • Clinical appeal summarization
  • Workflow optimization
• Build end-to-end ML pipelines: feature engineering, model training, hyperparameter tuning, validation, and monitoring.
• Research and apply advanced LLM techniques including embeddings, RAG (retrieval-augmented generation), prompt engineering, and document intelligence.
• Translate operational and product requirements into measurable model objectives, data specifications, and evaluation frameworks.
• Perform exploratory data analysis (EDA) to find patterns, anomalies, and business insights.
• Implement MLOps best practices (model versioning, reproducibility, feature stores, drift monitoring).
• Partner with Data Engineering to ensure feature availability, data quality, and scalable deployment using Azure Databricks.
• Collaborate with cross-functional teams to communicate insights and drive adoption of ML solutions.

Required Qualifications

• B.S./M.S. in Data Science, Computer Science, Statistics, Applied Mathematics, or related quantitative field.
• 8+ years hands-on experience building applied ML/AI models (healthcare experience strongly preferred).
• Strong proficiency in:
  • Databricks Python (pandas, scikit-learn, NumPy, TensorFlow, MLflow)
  • SQL
• Strong working knowledge of GitHub Copilot (or equivalent enterprise-approved AI coding assistant).
• Ability to thrive in a fast-paced environment with rapid prototyping and iterative delivery.

Preferred Qualifications

• Knowledge of US healthcare, ideally Revenue Cycle Management (RCM).
• Familiarity with healthcare EDI transactions such as 837 and 835.