• Serve as the team’s technical lead and mentor, guiding ICs through design, experimentation, and implementation while raising the technical bar. • Define and drive the Safety ML team’s technical strategy and roadmap. • Set the bar in technical reviews, including code, design, and architecture, ensuring the team builds scalable, robust, and high-quality ML systems. • Collaborate cross-functionally with product, data science, policy, legal, and engineering partners to align on safety goals and deliver effective solutions. • Tackle the most complex and high-impact challenges in safety, including adversarial abuse, harmful content detection, and evolving threat vectors. • Develop cutting-edge safety techniques, applying state-of-the-art ML to detect and prevent harm while staying ahead of emerging abuse patterns • Influence the company’s direction on safety, clearly communicating trade-offs and technical constraints to senior leadership and stakeholders.

• Help develop AI-based biomarkers that support the personalization of cancer therapy. • Work on multimodal deep-learning models that integrate large-scale pathology whole-slide images with diverse modalities to predict patient risk and response to therapy. • Contribute to developing robust, scalable solutions in this space. • Collaborate closely with fellow ML engineers, biostatisticians, scientific directors, and product partners to build, validate, and deploy models that ultimately impact patient care.

This description is a summary of our understanding of the job description. Click on 'Apply' button to find out more. Role Description We are seeking a talented ML Research Consultant to help us on our mission. You will work closely with our ML team to ideate, implement, and validate the machine learning classifiers and generative models that will ultimately improve our therapeutic design platform, translating functional activity data into breakthrough biologics. - Evaluate Sequence-Function Models: - Design, implement, and evaluate state-of-the-art machine learning models, including Protein Large Language Models (PLLMs), to predict functional activity of GPCR agonist antibodies. - Implement Generative Models: - Design, implement, and evaluate state-of-the-art generative models using Reinforcement Learning or Bayesian Optimization for de novo GPCR agonist antibody design. - Experimentation: - Design and execute rigorous ML experiments with clear hypotheses and documented outcomes and next steps. Qualifications - Bachelor’s or Master’s (5+ years of experience) or Ph. D. degree (2+ years of experience) in Bioinformatics, Biophysics, Computational Biology, or a related quantitative field. - Experience solving complex business problems in the life sciences using ML or statistical methods. - Excellent communication skills particularly in the context of cross-functional life science teams. - Understanding of modern deep learning architectures and optimization techniques as well as fundamentals of machine learning. - 2+ years of hands-on experience developing deep learning models using frameworks like PyTorch. - Strong proficiency in python and the scientific computing stack (NumPy, Pandas, scikit-learn). - Proficiency with software engineering best practices, version control (Git), and testing frameworks. - Experience with Transformer architectures, Recurrent Neural Networks (RNNs), or Graph Neural Networks (GNNs) applied to biological sequence or molecular/structural data. Preferred Qualifications - Theoretical foundations and practical experience with methods in computational structural biology, e.g., protein structure prediction, molecular dynamics, coarse-grained modeling, etc. - Experience with high-throughput screening data, particularly NGS analysis, flow cytometry, or other assays that link genotype to phenotype. - Direct experience with Protein Large Language Models (PLLMs), generative modeling, and/or Reinforcement Learning (RL) in a protein engineering context. - Background in developing models predicting antibody developability characteristics (e.g., stability, aggregation). - Background in molecular/cell biology, antibody discovery, bioinformatics. Benefits - Rate: 125-250 per hour depending on experience, up to 10 hours / week.

• Design, implement, and tune reinforcement learning algorithms to support decision making and behavior modeling in simulation environments. • Define and refine reward functions, state representations, and action spaces aligned to mission objectives. • Build and maintain training environments and experimentation pipelines for reinforcement learning agents. • Evaluate agent performance using rigorous metrics such as convergence, sample efficiency, robustness, and stability. • Analyze training outcomes and simulation results to recommend changes to models, scenarios, or learning approaches. • Collaborate closely with modeling and simulation engineers and software teams to integrate RL components into the baseline. • Ensure AI and ML components are transparent, testable, and maintainable within the broader system architecture. • Support test events, demonstrations, and technical reviews by clearly communicating AI and ML behavior and results.