Senior Research Engineer, Deep Genomics

• You are a research engineer who bridges the gap between fast-paced, experimental academic ML research and robust production systems
• You possess a solid mathematical foundation and a deep understanding of modern ML architectures, combined with the technical skills to write clean and highly optimized PyTorch code
• You thrive in ambiguity and are perfectly comfortable diving into messy research scripts to refactor, optimize, and build out the necessary scaffolding
• You understand the delicate balance between rapid iteration and technical debt, and you take pride in creating core tooling that empowers research teams to move fast without breaking everything
• Above all, you are a critical thinker and collaborative problem solver eager to push the boundaries of research
• Solid foundational grasp of linear algebra, calculus, and probability
• Strong understanding of modern machine learning/deep learning architectures and training dynamics
• High proficiency in PyTorch, including model building and basic optimization
• Strong general programming skills, with practical experience handling concurrency, threading, and memory management
• Demonstrated ability to debug software correctness and computational performance
• High tolerance for ambiguity and a willingness to work hands-on with unstructured research code

Desirable

• Develop and Optimize Core Tooling: Build and maintain the engineering infrastructure that allows the research team to iterate rapidly and safely
• Bridge Research and Engineering: Refactor and optimize experimental, script-like research code, adding necessary scaffolding and engineering rigor without stifling discovery
• Model Implementation: Implement, train, and evaluate modern deep learning architectures using PyTorch
• Testing and Debugging: Rigorously test and troubleshoot complex ML systems to ensure both software correctness and optimal computational efficiency
• Navigate Trade-offs: Continuously balance the need for research speed with the realities of technical debt, making pragmatic architectural decisions