Applied Data Scientist

Vi; Boston (MA); On-site; USD 120,000 - 180,000; Full time

Job summary

Vi Engage is seeking an Applied Data Scientist to manage 3–5 enterprise healthcare accounts. You will design pilots, integrate customer data (tokens, claims, EHR, marketing) into Vi's platform, and build automated pipelines that train, score, and deliver insights on a schedule. You remain the technical owner from onboarding to autopilot.

You will sit with customer teams to understand KPI needs and translate them into Vi platform requirements, shaping the roadmap with field learnings.

Qualifications

• Experimental design you can defend in business terms.
• Real data science depth: explain model workings and limits.
• Ability to ship production code and automate client deliverables with Python stack.
• Able to run client-facing sessions with health system stakeholders and travel onsite.

Responsibilities

• Own data integration pipelines from customer systems to Vi's data platform.
• Design pilots that tie model performance to business outcomes.
• Build per-customer automated production pipelines that train, score, and deliver insights.
• Serve as technical owner during onboarding and ongoing operations.
• Synthesize common patterns into platform requirements for faster deployments.

Skills

Python

ML engineering

PySpark

Pandas

Scikit-learn

Airflow

Tools

AWS S3

AWS Glue

AWS EMR

MWAA

SageMaker

Job description

Vi Engage puts predictive models into the daily operations of the largest health systems and health plans in the country — driving care navigation, specialty capture, and the workflows that follow from them. Applied Data Scientists are the people who turn a new customer into a running deployment on Vi's platform.

You will own 3–5 enterprise healthcare accounts at a time, end to end. For each one you are the data expert in the room: you design the pilot that proves value, integrate the customer's data (tokens, claims, EHR, marketing) into Vi's platform and build the automated pipelines that train, score, and deliver insights on a schedule. You stay the technical owner through operations and drive accounts to “autopilot”.

This is a hands-on-keyboard role with direct exposure to stakeholders. You are expected to sit with customer teams with expertise in analytics, marketing, and clinical operations to understand what KPIs they need to move, and how to leverage Vi’s capabilities to make it a reality. What you learn in the field becomes the product: you will spot the patterns across your accounts and turn them into the requirements that shape Vi's roadmap.

What You'll Own

• Data integration. Own the pipeline between customer data systems and Vi's data platform — ingestion, mapping, quality, and the judgment calls about what the data can and cannot support.
• Pilot design. Design studies that tie model performance to a business outcome the customer will underwrite. Power, controls, confounders, and a metric that survives scrutiny from a sophisticated internal analytics team.
• Automated production pipelines. Build per-customer pipelines that use Vi’s ML capabilities to train, score, and deliver insights without manual intervention — and keep them running.
• Customer ownership. Be the technical authority across onboarding and ongoing operations, including the ad-hoc data and reporting questions that come with a live account.
• Product leverage. Identify what's common across customers and synthesize it into high-leverage requirements for the platform, so the next deployment is faster than the last.

What We're Looking For

• Experimental design you can defend. You have designed and run studies where the result mattered commercially — and you can explain, to a skeptical stakeholder, why the design supports the conclusion. This is the capability we screen hardest on.
• Real data science depth. You can explain how a model works, evaluate it honestly, and say what it is not good for. You know when a simpler approach is the right answer.
• Python and ML engineering. Fluent in Python and the working stack — pandas, sklearn, airflow. You ship code to production that automates client deliverables.
• PySpark at scale. You have built distributed data and ML pipelines on Spark against large, messy datasets.
• Client-facing command. You can run a working session with clinical, IT, and analytics leaders at a major health system or health plan on your own — translating between their business problem and what the data will actually support, and holding the line when the answer isn't what they hoped. You will occasionally travel onsite with clients.
• Ownership instinct. You treat your accounts as yours: you find the problem before the customer does and you fix it.

Nice To Have

• Healthcare or life sciences domain knowledge — claims, EHR, HL7/FHIR, lab data, or population health analytics
• Experience designing and optimizing the targeting of marketing and engagement campaigns
• AWS data infrastructure: S3, Glue, EMR, MWAA, SageMaker
• Familiarity with HIPAA and healthcare compliance and data governance frameworks
• Experience taking a product from its first customer deployment to a repeatable one

What This Role Is Not

This is an applied, in-production role. It is not a research position — the work is measured by deployments that run and outcomes customers can point to, not by novelty. And it is not a delivery or engagement management role: you will be architecting and writing the pipelines yourself, not coordinating someone else who does.