Senior Data Scientist

• Vi is an enterprise AI platform for health enterprises - healthcare, biopharma, and wellness. We deploy agentic AI and predictive models into production environments where the output drives the next best actions for patients, care teams, and operations to deliver ROI and improve health outcomes.
• The Vi Platform is the shared layer under every Vi application. It turns data, models, agents and compliance into services that every product uses to onboard clients, reach patients and measure results, so a capability is built once and reused everywhere. We are looking for a senior data scientist to take those services from idea to production, and to keep them running as they grow.
• You will be part of the Vi Platform team, reporting to the VP of R&D. This is a hands-on IC role at the crossroads of data science, software engineering and cloud infrastructure: you design the service, you build every layer of it, you ship it, you measure what it does for the products that use it, and you maintain and extend it with the features its users need.

Responsibilities

• Train and evaluate models on tabular data: Turn patient- and member-level tabular data - claims, clinical, engagement and vendor feeds - into features, labels and models (gradient boosting, tabular foundation models, embeddings), with an evaluation harness that defines what “good” means before a product feels a change.
• Own services end to end: From a one-page design through API, SDK, UI and infrastructure to production, built to the platform's service standards: tested at every level, auto-deployed on merge, with live status and cost per job.
• Build data services at scale: Spark and EMR jobs over the lakehouse, orchestrated in Airflow, that resolve identities, join vendor feeds and score millions of records per run, and that get faster and cheaper with every release.
• Put models and LLMs into production: Foundation model, embedding, classification and agent services, each with an evaluation harness behind it, so a model change is measured before a product feels it.
• Build the full stack: Backend services, web consoles and SDKs that an engineer, a product team or an agent can use self-serve, with a user guide for people and a playbook for agents.
• Run what you ship: Monitoring, alerting, retries and cost per job. You are the first responder when your service is red, and you announce every release to its users with measured numbers.
• Work AI-native: AI coding agents are part of how we build every day. You drive them, review their output critically, and build services that agents consume as easily as people do.

Requirements

• 4+ years of hands-on data science experience, with a track record of training, evaluating and shipping models that run in production as a hands-on IC.
• Data science and ML in production: Building and evaluating models (tabular, embeddings, LLM-based), deciding what "good" means for a service, and measuring it.