SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Vi Engage deploys predictive models into the operations of major health systems and health plans to drive care navigation, specialty capture, and downstream workflows. As Staff Data Scientist, you will own the ML pipelines and modeling engine that power these deployments across the entire customer base.
You will architect and build config-driven machinery that transforms longitudinal claims, EHR, lab, and behavioral intent data into patient risk and enrollment propensity predictions. Rather than building one-off models for individual customers, you will create a generalizable platform with automatic feature engineering and model selection that reproduces lift across dozens of deployments without per-customer engineering work.
Key responsibilities include:
- Designing and maintaining ML pipelines built for scale, handling large longitudinal datasets with config-driven training, scoring, and delivery
- Implementing mass customization through automatic feature engineering and model selection to produce customer-specific models without custom engineering
- Identifying modeling improvements that generalize across all customer deployments, measuring success by cross-customer lift
- Building reusable components that Forward Deployed Data Scientists compose into production deployments, accelerating time-to-value for new customers
- Productizing pilots and proofs of concept into durable capabilities that survive multiple customer implementations
- Setting the modeling standard and best practices that the broader data science team builds against
This is an applied, production-focused role measured by pipelines that run reliably and models that hold up across customers, not by research novelty. You will be hands-on architecting and building with the team, not managing people. You may occasionally interface with design partner clients but are not customer-facing day-to-day.
REQUIREMENTS:
- Demonstrated experience shipping uplift, survival, or propensity models in production whose output changed organizational spending or reach decisions (screened hardest on this capability)
- Proven track record taking pilots and proofs of concept to repeatable, scalable products that survived multiple customer implementations
- Deep real-world data science expertise: segmentation, campaign optimization, uplift estimation strategy, model evaluation, and judgment to recommend simpler approaches when appropriate
- Fluent in Python and ML engineering stack: pandas, sklearn, PySpark, Airflow
- Hands-on MLOps experience: model tracking and deployment tooling (mlflow, SageMaker, or equivalent)
- AWS and cloud proficiency: able to deploy models at scale without handing off to dedicated engineers; working knowledge of S3, Glue, EMR, MWAA, SageMaker
NICE TO HAVE:
- Healthcare or life sciences domain knowledge (claims, EHR, HL7/FHIR, lab data, population health analytics)
- HIPAA and healthcare compliance and data governance framework familiarity
- Experience designing pilots and efficacy studies that tie model performance to business outcomes
- Experience building internal platforms or frameworks that other engineers build on top of
