Note: The job is a remote job and is open to candidates in USA. Careforth is seeking a Data Scientist II to join its AI/ML organization and develop production-ready solutions for risk intelligence, NLP, and AI enablement products. The role focuses on building and deploying machine learning, NLP, LLM, speech, and MLOps pipelines while ensuring explainability, reliability, and HIPAA compliance.
Responsibilities
- Build and deploy supervised classification models (XGBoost, LightGBM, regularized regression) for clinical risk stratification and caregiver engagement scoring, including feature engineering, class imbalance handling, model calibration, and threshold optimization
- Develop NLP pipelines for clinical text extraction from nurse notes and operational data — including named entity recognition, negation detection, relation extraction, and document classification using spaCy, scispaCy, and MedSpaCy
- Implement and maintain LLM-powered tools including RAG architectures, prompt engineering, embedding models and vector search, hallucination detection, and response evaluation and guardrails for clinical and caregiver-facing use cases
- Support speech and audio ML pipelines — ASR fine-tuning, speaker diarization, sentiment and emotion detection, and keyword detection — for real-time and ambient listening applications
- Contribute to composite risk scoring and ensemble model fusion, including SHAP-based explainability, model calibration, and uncertainty quantification supporting care team and payer-facing outputs
- Apply model explainability (SHAP, LIME), monitor for data drift and bias, and uphold HIPAA-compliant data handling standards throughout the ML lifecycle
- Maintain MLOps standards: implement automated testing, CI/CD for ML, model versioning via MLflow, and participate in experiment tracking and model registry governance in Databricks
- Collaborate with Data Engineering on feature store integration, scoring pipeline design, and structured output schemas; participate in technical design reviews and provide peer feedback
- Perform other duties and special projects as assigned
Skills
- • Bachelor's or Master's degree in Computer Science, Data Science, Computational Linguistics, Statistics, or a related quantitative field
- • 4–6 years of applied ML experience, with at least 2 years in production NLP or LLM development
- • Demonstrated ability to take models from experimentation through deployment and ongoing monitoring
- • familiarity with HIPAA data handling required
- • Strong Python: pandas, NumPy, scikit-learn; deep learning frameworks (PyTorch and/or TensorFlow)
- • Proficiency in gradient boosting methods (XGBoost, LightGBM), classification modeling, and model evaluation including cross-validation, precision-recall optimization, and SHAP-based explainability
- • Hands-on NLP experience with spaCy, scispaCy, MedSpaCy, Hugging Face Transformers (BERT, GPT), and LangChain; experience with LLM fine-tuning, RAG, and vector databases
- • Familiarity with speech ML tools: ASR models (Whisper, Azure Speech), speaker diarization, and sentiment and emotion detection
- • Experience with Databricks (Delta Lake, MLflow, Spark) and AWS (S3, SageMaker, Bedrock); working knowledge of Docker and CI/CD for ML pipelines
- • Solid SQL skills; ability to work across data lakehouse and relational environments
- • Clear and precise communicator; able to document and explain model behavior to both technical and non-technical stakeholders
- • Collaborative team contributor; receptive to feedback, proactive in design discussions, and committed to shared engineering standards
- • Intellectually curious and self-directed; actively keeps pace with developments in applied ML and clinical AI
- • Experience in healthcare, managed care, or regulated data environments preferred
Benefits
- Flexible schedules
- Remote-first culture
- A nationally recognized wellness program
Company Overview