Staff AI Scientist (Foundation Models)

Verge Genomics • San Francisco, CA

5d

About The Position

Who We Are Verge is transforming drug discovery by using artificial intelligence and proprietary human data to solve the biggest driver of rising drug costs: high clinical failure rates. To achieve this, we have built one of the field’s largest corpuses of multi-modal patient molecular and clinical data, sourced directly from human tissue. Our team of engineers, neuroscientists, and biologists have so far delivered two drugs to clinic, discovered 282 new targets, and signed commercial partnerships worth in excess of $1.6B with Eli Lily and AstraZeneca. Your Mission Reporting directly to the Head of Product & Engineering, and working closely with Verge’s engineering and computational biology teams, the Staff AI Scientist (Foundation Models) will be responsible for leading the design and development of next-generation, neuroscience foundation models powered by Verge proprietary multi-modal patient datsets–empowering Verge’s customers and partners with state of the art performance in disease-specific tasks, and working in partnership with leading frontier AI companies to move the field forwards. Your 12 Month Outcomes Assess and nominate one candidate from several potential translational applications for Verge’s first foundation model, Design benchmarks and evaluations to validate the model’s breakthrough performance throughout each stage of its development, Deliver a state of the art model, scoping and hitting key milestones–starting with a proof of concept and culminating in authoring a paper validating its performance through robust experiments. You Will… …take a hands on approach to rapidly building a proof of concept that demonstrates the potential for state of the art performance, …lead additional collaborators across data curation, infrastructure, and computational biology as you achieve milestones and scale up Verge’s investment in the model, …translate between biological domain knowledge and AI objectives, both within Verge and between Verge and its AI and industry partners.

Requirements

• PhD in computer science, computational biology, AI/ML, applied statistics, biophysics, or related field, or MS in related field with exceptional professional experience
• Demonstrated experience training and evaluating modern ML methods (e.g. transformers, diffusion-based generative models, graph neural networks, sequence models)
• Proven ability to contribute at every stage of the design and delivery of foundation models yielding novel scientific discoveries
• Demonstrable ability to design and run non-trivial experiments: controlling for confounders, building robust baselines, thorough error analysis, etc.
• Exceptional grasp of modern AI/ML frameworks and methods (PyTorch, JAX, Tensorflow, diffusion-based generative models, graph neural networks, etc.)
• Ability to thrive in uncertainty with frequently changing priorities
• Deep alignment with our values
• A passion for making an impact on patients

Nice To Haves

• Experience with large-scale data (e.g. 100B+ tokens) or distributed training
• Background in computational biology, genomics, or a related field
• Experience working with multimodal foundation models
• First-hand experience building and evaluating transformer-based models using biological data (e.g. scGPT, Geneformer, ESM, rBio, ProtBERT)
• Technical expertise in model optimization (e.g. FSDP/ZeRO, quantization, compilation, custom kernels)
• Experience with approaches to federated learning (e.g. NVFLARE)

Responsibilities

• take a hands on approach to rapidly building a proof of concept that demonstrates the potential for state of the art performance
• lead additional collaborators across data curation, infrastructure, and computational biology as you achieve milestones and scale up Verge’s investment in the model
• translate between biological domain knowledge and AI objectives, both within Verge and between Verge and its AI and industry partners

What This Job Offers

• Job Type: Full-time
• Career Level: Mid Level
• Education Level: Ph.D. or professional degree