Head of Data Science - Claims Intelligence

Company Overview: Claims Intelligence, a Vantora (formerly UP.Labs) portfolio company, is hiring a hands-on Head of Data Science. Claims Intelligence is building an AI-powered claims and litigation intelligence platform that helps commercial trucking carriers move from reactive defense to proactive risk management.

Commercial auto claims are a $32B+ annual problem, and the cost keeps growing. The moments that matter most in a claim happen in its first 72 hours. Once plaintiff counsel gets involved, settlement costs can multiply several times over. Today, carriers’ claims teams are working from disconnected tools, examiners carry far more open claims than the industry standard, and critical judgment lives in the heads of a few senior staff. Claims Intelligence is a real-time decision layer embedded directly in the claims workflow. It surfaces the right guidance at intake, triages and routes claims automatically, and gives examiners data-driven settlement and fraud intelligence.

As the first data science leader, you will own the modeling foundation of the company end-to-end. That includes the claims ontology, risk and severity models, settlement intelligence, and the long-term vision for a cross-carrier data asset. You’ll stay deeply hands-on, building and shipping models yourself, while serving as the company’s thought leader on claims analytics for engineering, product, customers, and leadership.

You’ll also have access to the Vantora ecosystem, a world-class team across product, engineering, design, analytics, marketing, legal, talent, finance, and venture capital, to help accelerate execution.

In This Role, You Will:

• Build and own the core modeling layer of the platform, including:
  • Bodily injury risk-flagging and severity prediction at first notice of loss (FNOL)
  • Escalation modeling (predicting which claims are likely to progress to bodily injury or litigation)
  • Settlement banding by jurisdiction, injury type, and historical outcomes
  • Fraud and payment identity analysis
• Design the claims ontology and data model that turns messy, inconsistent claims data into a single source of truth.
• Apply rigorous statistical methods to small, heavy-tailed, and censored datasets (e.g., GLMs, survival/time-to-event models, hierarchical and Bayesian approaches), using ML and LLMs where they add real value.
• Build models that are explainable and trusted by claims examiners, risk managers, and executives, not black boxes.
• Identify and bring in external data sources (e.g., jurisdictional, verdict, and litigation data) that meaningfully improve outcomes beyond what any single carrier’s data can show.
• Help define the architecture and governance for an anonymized, cross-carrier consortium dataset, the company’s long-term data moat.
• Partner closely with Engineering and Product to translate models into production features embedded in live claims workflows.
• Translate complex statistical concepts for technical and non-technical audiences, including engineers, product leaders, design partners, and the CEO.
• Lay the foundation for a future data science team, with the opportunity to hire and lead as the company scales.

Who You Are:

• A hands-on data science leader who builds and ships, not just manages.
• Deeply experienced in insurance, claims, or risk modeling, and comfortable with the realities of claims data.
• A strong statistician at heart: you know when a simple, well-specified model beats a complex one, and you can defend that choice.
• An open, creative thinker who looks beyond what the model says to ask what data and interventions would actually move the needle.
• A natural explainer and thought leader who can bring a non-insurance team up to speed quickly.
• Energized by ambiguity and the pace of a 0→1 startup.

You Should Have:

• 7+ years of experience in data science, statistical modeling, or quantitative analytics, with meaningful time in insurtech, insurance, claims, or another insurance-adjacent domain (P&C, workers’ comp, disability, life, or health).
• An advanced degree (MS or PhD) in Statistics, Actuarial Science, Mathematics, or a related quantitative field. Strong computer science backgrounds with deep insurance modeling experience will also be considered.
• Expert-level Python (pandas, NumPy, SciPy, scikit-learn, statsmodels) and fluency in R.
• Proven experience taking models from research to production in a product environment.
• Experience with claims severity, reserving, litigation propensity, or fraud modeling is a strong plus.
• Experience in a startup or high-growth environment is strongly preferred.
• Familiarity with Google Cloud Platform (BigQuery, Vertex AI) is a plus.
• Background in transportation, logistics, or trucking is not required.

Location: New York City preferred. Exceptional remote candidates will be considered.

Why Join:

• Solve a high-stakes, high-dollar problem where better decisions in the first 72 hours of a claim directly reduce settlement costs.
• Build a category-defining platform in a whitespace no existing player occupies: carrier-side, bodily-injury focused, and embedded in the examiner workflow.
• Own the modeling foundation from day one, with real influence over the product and the company’s data strategy.
• Work with leading trucking carriers already engaged as design partners.
• Meaningful equity ownership in a high-growth venture with enterprise-scale potential.

About Vantora: Vantora builds high-growth technology startups that enable faster, cleaner, and safer movement of people and goods.

Our platform is unique in three ways:

• Risk: We reward our team and partners with meaningful equity.
• Technology: We build and launch scalable technology products from day one.
• Industry Focus: We stay deeply focused on the underlying fabric of mobility and logistics.