Despite the emergence of new modalities and drug development technologies, the cost and time to produce new therapies has changed little, and failure rates remain high. Xaira aims to change that with a systematic, AI‑driven approach that tackles three pervasive bottlenecks—choosing the right targets, designing the right molecules, and matching the right patients—by running as much work as possible in silico and using high‑dimensional causal datasets to train “virtual cell” foundation models. The company is initially focusing on high‑value, historically undruggable targets and ultimately on building a pipeline of differentiated biologics. We spoke with Marc Tessier‑Lavigne, co‑founder and CEO of Xaira, about applying end‑to‑end AI across target discovery, molecular design, and patient stratification, the company’s more than $1 billion in funding, and how it seeks to enable a new generation of scientists fluent in both AI and biology.