ML Researcher

About the Company

Axiom is building a compounding ecosystem to replace animal testing and reshape how clinical trials are run. By building world-class datasets and advancing ML research, Axiom develops models that predict human drug outcomes more accurately than animal models, currently focusing on drug-induced liver injury for top pharma companies.

Responsibilities

• Define end-to-end ML and agent systems spanning data generation, cleaning, feature extraction, and deployment
• Build novel models learning relationships between chemistry, biological response, dose, and human toxicity
• Train large multimodal models on chemical structures, cellular images, transcriptomics, and proteomics
• Develop foundation models and representation-learning systems for biological images and molecules
• Architect models to predict human toxicity based on dose, potency, and chemical structure
• Collaborate with computational biologists, chemists, and wet-lab teams to design experiments
• Own the research-to-product loop from prototyping to shipping models to customers

Requirements

• Exceptional machine learning ability demonstrated through industry, academia, or open source work
• Deeply technical with proficiency in PyTorch and debugging training runs
• Ability to work with noisy, multimodal, and sparse biological data
• Capability to move between research ideas and production systems
• Strong engineering ability and interest in scaling inference and building real systems
• Curiosity to learn biology, chemistry, toxicology, and pharmacology

Preferred Qualifications

• Experience with JAX, TensorFlow, or other deep learning frameworks
• Expertise in representation learning, contrastive learning, or self-supervised learning
• Experience with computer vision for biological imaging or high-content screening
• Knowledge of LLMs, agents, retrieval, and reasoning systems
• Familiarity with large-scale distributed training and GPU infrastructure