Principal ML Scientist, Multimodal Biological Reasoning
$216,000–$297,000 year
On-siteCambridge, Massachusetts, United States
Job Summary
Lead the roadmap for multimodal biological reasoning engines, guiding architecture from data exploration through benchmarking and agentic integration. Translate high-value use cases from Flagship teams and portfolio companies into technical requirements, defining success criteria and creating feedback loops for autonomous science workflows. Build and integrate reasoning engines capable of mechanism-of-action reasoning, target discovery, and pathway analysis into the AI Scientist platform. Source data assets across genomics, transcriptomics, and protein modalities to drive model development that prioritizes biological relevance over pure benchmark performance. Communicate strategic direction and scientific results to stakeholders and public forums while originating multiple projects in this space.
Required Qualifications
- PhD, MS, or equivalent experience in computational biology, machine learning, bioengineering, computer science, systems biology, quantitative biology, or a related field
- Experience in biotech, pharma, AI-for-science, AI drug discovery, venture creation, or a platform organization serving multiple scientific programs
- Experience deploying AI/ML for biology systems into scientific workflows at enterprise scale
- Experience with LLMs, biological foundation models, protein language models, genomic foundation models, scientific agents, or AI discovery platforms
- Experience working across multiple internal and external customers, therapeutic programs, or discovery teams
Desired Qualifications
- Industry-leading expertise in modern LLMs, multimodal modeling, representation learning, fine-tuning, post-training, benchmarking, and ML systems
- Deep experience in computational biology, AI-for-biology, AI-enabled drug discovery, translational data science, biological foundation models, or scientific discovery platforms
- Demonstrated ability to work with scientists or biotech teams to scope high-value use cases, identify data assets, define success criteria, and translate discovery needs into technical execution plans
- Working knowledge of biological data modalities such as genomics, transcriptomics, perturbation data, protein sequence, protein structure, pathways, imaging, pathology, or time-series biological data
- Strong judgment around mechanism-of-action reasoning, target discovery, perturbation biology, scientific credibility, model limitations, hallucination risk, interpretability, and validation
- Experience leading small, high-caliber technical teams through ambiguous scientific or product problems
- Ability to communicate clearly with ML researchers, data engineers, portfolio-company scientists, executives, and venture creation leaders
- Comfort operating in an entrepreneurial environment where the goal is not only to build a model, but to create a new capability that can reshape company creation and scientific discovery
Hiring someone like this?
Get your role in front of qualified candidates on Sorce.