Applied Scientist, Oncology Foundation Model

Pathos is developing an AI-driven biotech platform centered on oncology. The company operates with minimal hierarchy, utilizing small expert teams (2-4 people) who direct AI agents to accomplish work previously requiring dozens of staff members.

Key Responsibilities

• Lead development and training of specialized language models for oncology
• Design strategies for managing oncology-specific datasets at scale
• Implement alignment techniques including RLHF and supervised fine-tuning
• Create evaluation frameworks assessing model performance and clinical relevance
• Conduct biomedical LLM research
• Publish in top-tier conferences and communicate findings broadly
• Collaborate with oncologists and clinical teams
• Mentor junior scientists

Required Qualifications

• PhD in Computer Science, Machine Learning, AI, or related field (or MS + 8+ years equivalent experience)
• 5+ years hands-on deep learning experience
• Proven expertise in pretraining and post-training large language models
• Strong publication record at NeurIPS, ICML, ICLR, ACL, or EMNLP
• Proficiency in Python and deep learning frameworks (PyTorch or TensorFlow/JAX)
• Experience with distributed training and large-scale model infrastructure

Preferred Qualifications

• Experience applying LLMs to biomedical or healthcare domains
• Background in computational biology or medical informatics
• Knowledge of oncology, clinical workflows, or cancer biology
• Experience with RAG or knowledge grounding
• Model safety and responsible AI expertise
• Track record translating research into production systems