Concr is a mission-driven techbio company leveraging statistical learning and advanced modelling adopted from astrophysics to answer essential questions in oncology. By empowering cross-talk and iterative learning between disparate data sources, Concr's powerful predictive engine can accurately identify molecular features of patient response, also known as biomarkers. Over 90% of novel cancer drugs fail early in clinical trials, leading to more and more data being generated, which will yield limited value unless it can be interlinked to generate meaningful biological insights into individualised patient response to therapy. At Concr, we use established cosmology methods with decades of use to integrate limited and sparse data optimally, even with missing data points. By combining cutting-edge machine learning techniques with a deep scientific understanding of tumour progression, we can accurately predict cancer evolution in response to treatment. Our cloud-based platform, FarrSight™, allows users to store, manage and analyse their data in the same place, effortlessly generating actionable insights for cost-effective drug development and patient treatment. This unique technology enables our partners to identify and develop biomarkers for their drugs, with 300x less pre-clinical and 7x less clinical data compared to other approaches.