Proven ability to deliver high-quality, maintainable code at a rapid pace.
Experience designing and architecting complex systems from the ground up.
Adaptive mindset with a history of quickly mastering new and complex fields.
Demonstrated success in solving challenging engineering problems, including distributed systems and kernel optimization.
Responsibilities
Define and architect a high-performance infrastructure for model training and simulations.
Design and optimize custom reinforcement learning environments based on neuroscience principles.
Establish standards for efficient, production-grade code to enhance research speed.
Transform research prototypes into scalable production systems.
Identify and resolve technical bottlenecks to support research goals.
Benefits
Access to computational resources typically unavailable in traditional foundations.
Opportunity to work in a high-velocity startup-like environment.
Freedom from funding pressures and profit motives to focus on ambitious scientific goals.
Collaboration with top-tier scientists and engineers from leading academic institutions.
Full Job Description
About Astera:
Astera is a private foundation with a $2.5B endowment on a mission to steer science and technology toward an abundant future for all. Unlike traditional foundations, we operate like a high-velocity startup with unprecedented access to computational resources and complete freedom from funding pressures or profit motives. This allows us to focus on ambitious goals and attract incredibly creative scientists and engineers from leading academic institutions and from frontier AI labs.
Obelisk is our large-scale AI research program, pursuing a neuroscience-informed approach to engineering AGI. This is not yet-another-lab scaling LLMs in a hope of achieving general intelligence. We are integrating neuroscience, AI, and bioengineering to understand and digitally model the architecture of the human brain.
Position Summary:
We are looking for an ambitious Staff Software Engineer to help us reverse-engineer the human brain.
This role spans a broad spectrum of activities:
Define and architect high-performance infrastructure for training large-scale models and running complex simulations.
Design and optimize custom RL environments informed by neuroscience.
Set the standard for clean, fast, production-grade code that boosts our research velocity.
Turn research prototypes into scalable production systems.
Tackle systemic technical bottlenecks, from low-level optimization to high-level strategy, to unblock the research team.
Qualifications:
High-Velocity Craftsmanship: You set the bar for clean, maintainable, and well-tested code at a startup pace. You know how to balance engineering rigor with the need to iterate quickly on research prototypes.
System Architecture Intuition: You have proven success architecting complex systems from scratch. You understand the long-term trade-offs, and you can design infrastructure that survives the chaos of research.
Extreme Adaptability: You might be an expert in distributed systems today, but you are ready to become a novice in neuroscience tomorrow. You have a track record of diving into complex domains, mastering them quickly, and guiding others through them.
Mastery of Hard Problems: You have a track record of leading the resolution of difficult engineering challenges, whether it is massive scale distributed systems or kernel-level optimization. You do not shy away from what your peers consider hardcore.
Location & Visa:
This role is in-person in Emeryville, CA.
Visa sponsorship may be available for qualified candidates.
