Trial Principal, Statistician, Roivant Health

Position Summary

The Principal, Statistician will own the statistical and quantitative work that connects clinical evidence to the most important decisions Roivant makes across its pipeline. This is not a strategy role — this is a figure‑things‑out and get‑things‑done role in a high‑stakes environment with substantial ownership. We expect you to get to ground truth on the hardest clinical and scientific questions the company faces, working directly in trial datasets and translating rigorous analysis into recommendations that leadership acts on.

The Principal will pressure‑test the clinical assumptions underpinning key pipeline and commercial decisions, build the statistical frameworks needed to evaluate evidence across assets, and ensure Roivant has a clear‑eyed view of what its data actually shows. The scope is intentionally broad and is well‑suited to a high‑potential individual earlier in their career with exceptional upside and a strong desire to take on responsibility.

• Work directly in ongoing and completed clinical trial datasets (CSRs, patient‑level data, interim readouts) to identify efficacy signals, safety patterns, and subgroup effects that inform go/no‑go and commercialization decisions
• Build rigorous statistical analyses from scratch: mixed‑effects models, longitudinal models, Bayesian frameworks, subgroup analyses, and sensitivity analyses across Roivant's pipeline assets
• Monitor emerging safety signals and unexpected findings in real time across ongoing studies; surface and contextualize them for leadership before they become surprises
• Pressure‑test clinical assumptions: challenge endpoint choices, scrutinize statistical analysis plans, and flag where the data does not support the narrative
• Execute focused 1‑5 day analytical sprints on the highest‑priority questions from Roivant and Vant leadership
• Develop a deep, differentiated read of the clinical evidence base for Roivant's near‑commercial assets: what the data shows, what it does not, and where the science is genuinely uncertain
• Analyze competitors' trial data, published and unpublished, with the same rigor applied to Roivant's own: endpoints, patient populations, effect sizes, and durability
• Synthesize findings from literature, registries, claims, and proprietary datasets into coherent evidence narratives that hold up under scrutiny

• Drive selection and integration of the data sources needed to support portfolio decisions, including claims, labs, clinical trial data, and genomics
• Help build shared analytical infrastructure across Roivant and Vant companies
• Identify opportunities to apply consistent statistical frameworks across assets and therapeutic areas

• Exceptional academic background from a top‑tier program in statistics, mathematics, computer science, or a related quantitative field
• Deep command of statistical and computational methods: mixed‑effects models, longitudinal analyses, Bayesian approaches, machine learning, and the judgment to know which tool fits the problem
• Demonstrated interest in life sciences, biology, or health. While you do not need a biology degree, you are genuinely curious about how medicines work and what clinical data means
• Hands‑on experience working with large, complex datasets; you have built analyses from scratch, not just run existing pipelines
• High capacity for learning; you pick up new domains quickly and are comfortable operating in areas where you are not yet an expert
• Track record of translating rigorous analysis into decisions; your numbers have changed something
• Able to communicate complex quantitative findings clearly to non‑technical audiences, including CEO‑level stakeholders
• Intensely curious, resourceful, and motivated; you run toward hard, ambiguous problem

Roivant is preparing to become the next breakout commercial‑stage biotech, with the commercialization of brepocitinib, IMVT‑1402, and mosliciguat. Getting those launches right requires an unusually rigorous understanding of the clinical evidence that defines each market: what the data actually shows, not what anyone wishes it…