Position SummaryMissionHire a founding technical leader to serve as the CEO's technical co-pilot; the person who co-architects Radial's scientific and technical direction, builds the AI and computational infrastructure across programs, and sets the technical bar through their own hands-on work. This is a building role: the Head of Bio AI will personally drive early prototypes, make critical architecture decisions, and help determine what Radial builds, how it builds it, and the standards to which it holds itself.
Key Outcomes (12-24 Months)1. Technical Architecture & Program Leadership- Establish AI and computational architecture across current and future programs, beginning with DiffUSE and its expansion across multiple structural biology data modalities.
- Shape which foundational bottlenecks Radial addresses and how they become durable technical systems.
- Help make build-vs-fund-vs-partner decisions across Radial's program portfolio.
2. Hands-On Building- Conceive, architect, and directly prototype early modeling systems and core infrastructure.
- Design scalable training workflows, data pipelines, validation frameworks, and evaluation benchmarks.
- Evolve capabilities from early prototyping to production-grade, open platforms the scientific community can build on.
3. Team & Culture- Build a high-caliber AI and engineering team, starting small and scaling deliberately.
- Set the technical bar through hands-on contribution and architectural authorship.
- Establish a culture of technical rigor, intellectual honesty, and disciplined experimentation.
4. Ecosystem & Impact- Engage leading scientific collaborators and build technically rigorous partnerships.
- Represent Radial's technical vision within the AI and life sciences communities.
- Ensure outputs are structured and shared as durable public goods.
CompetenciesFunctional Expertise- Deep expertise in modern ML: large-scale training, representation learning, generative modeling (diffusion, transformers, foundation models).
- Track record of translating ideas into working, scalable systems.
- Understanding of the interplay between machine learning and physics-based modeling
- Direct experience formulating and solving inverse problems, including familiarity with ill-posedness, regularization strategies, and the trade-offs between learned and model-based reconstruction approaches
- Experience building and iterating on ML pipelines for large-scale, data-intensive problems, including efficient data ingestion, preprocessing of high-dimensional inputs, and training workflows that scale across distributed compute resources
- Systems-level thinking across data generation, modeling, infrastructure, and experimentation.
- Experience designing technical platforms intended to endure and compound over time.
- Scientific range beyond a single modality can engage across problem domains.
Leadership Attributes- Has managed small technical teams; knows how to attract and retain exceptional talent.
- Hands-on builder who sets the standard through their own work.
- Deep familiarity with the frontier AI x Bio landscape.
- Operates effectively in ambiguity; creates clarity through disciplined technical judgment.
Cultural Alignment- Motivated by the chance to build something that doesn't exist yet.
- Committed to open science and durable public infrastructure.
- Thrives in early-stage, open-ended environments blending AI and biology.
LocationThis role is hybrid, based in Emeryville, CA. The CEO and most of the team are on-site at least 3 days per week.
CompensationThe successful candidate will receive a competitive compensation package commensurate with their experience.
As a nonprofit organization, Astera is exempt from the H-1B cap; visa sponsorship may be available for qualified candidates.
Additional Information- Astera Institute Mission & Vision
- Astera Institute Overview
- Astera Institute Team
- Sharpening Astera's Focus
- Dileep George Joins Astera
- About the DiffUSE Project
- Seemay's Blog: From Systems Operators to Systems Architects
- DiffUSE January 2026 Retreat
- DiffUSE Project LinkedIn
