Founding Research Scientist, Biological Foundation Models

The CompanyA venture-backed technology company developing machine-learning systems for biological research and drug discovery. The team is building large-scale models that integrate diverse biological datasets to better understand cellular behavior, disease mechanisms, and therapeutic response.The company works closely with scientific and industry collaborators on applications spanning drug discovery, biological modeling, translational research, and predictive analytics.The RoleWe are looking for our Founding Research Scientist, Biological Foundation Models. This will be an early and highly autonomous research role, with ownership across model development, experimentation, evaluation, and scientific delivery.You will take complex biological questions and translate them into well-defined machine-learning problems, datasets, experiments, and models. You will work closely with company leadership and have substantial influence over research direction and modeling strategy.The ideal candidate combines strong experience training large-scale models with meaningful hands-on experience working with biological data.ResponsibilitiesLead machine-learning research projects from problem formulation through experimentation and deliveryDesign, train, fine-tune, and evaluate models using biological and multimodal datasetsWork with data including single-cell, spatial, imaging, perturbation, multi-omic, and other high-dimensional biological measurementsDevelop models for biological state, response prediction, perturbation modeling, and related applicationsDesign rigorous evaluation frameworks covering generalization, robustness, and biological validityCollaborate with experimental scientists to determine useful controls, validation strategies, and additional data requirementsTranslate research results into reusable modeling capabilities and practical workflowsPresent technical findings and modeling decisions to both technical and scientific audiencesPartner with engineering teams to make research systems scalable and reliableRequirementsDemonstrated experience training or substantially improving large machine-learning modelsStrong understanding of representation learning, model architecture, optimization, and evaluationExperience making decisions around pretraining, fine-tuning, and other model adaptation techniquesStrong Python and PyTorch or JAX skillsExperience with multi-GPU or distributed model trainingTrack record of models or research systems being used by other researchers, teams, or customersHands-on experience with biological datasets such as single-cell data, perturbation screens, spatial data, imaging, transcriptomics, or related modalitiesPractical understanding of biological noise, batch effects, confounding, controls, and experimental variabilityAbility to evaluate whether a model's apparent performance represents meaningful biological signalFamiliarity with current research in biological foundation models, virtual cells, perturbation modeling, or adjacent areasAbility to communicate effectively with computational and experimental biologistsWorking StyleComfortable owning an ambiguous research problem end-to-endEnjoys working in a small, rapidly evolving environmentStrong written and verbal communication skillsAble to move between research exploration and practical implementationComfortable collaborating directly with scientific leadership and external research teamsNice to HaveExperience training large vision, multimodal, protein, or other foundation modelsExperience applying machine learning to microscopy, histopathology, spatial biology, or high-content imagingExperience developing or benchmarking perturbation or biological-state modelsExperience working with pharmaceutical, biotechnology, or academic research organizationsPhD or equivalent research experience in machine learning, computational biology, or a related quantitative disciplineExperience with scientific agents, LLM post-training, tool use, or multimodal reasoningTo apply please reach out to Sam Shinner at Discover International.