Principal AI Scientist, Autonomous Discovery

Description

The company is building an autonomous system that searches large scientific and computational spaces. A core problem is designing algorithms that generate, score, refine and recombine candidate solutions across many iterative rounds, improving results over time while still exploring new directions.

The team is hiring a Senior Principal Scientist to design and build the next generation of evolutionary algorithms.

You will build methods where populations of models, hypotheses, programs or scientific representations evolve through repeated search, selection, variation and evaluation. The work draws on evolutionary computation, Bayesian inference, reinforcement learning, optimization and automated scientific discovery. You will shape how the system searches huge solution spaces, trading exploration against exploitation, keeping diversity alive and surfacing approaches that conventional optimization would likely miss. This is a senior individual contributor role with wide scientific freedom that directly steers the technical direction of the autonomous-science platform.

Responsibilities

• Design and implement algorithms that generate and evolve large populations of candidate solutions.
• Create principled ways to keep populations diverse, so the search finds truly distinct candidates instead of minor variations of one answer.
• Build Bayesian and probabilistic frameworks that evaluate, compare and rank candidates using mixed kinds of evidence, including estimates of uncertainty and redundancy across the population.
• Define and test quantitative measures of diversity and quality for a population.
• Work with research and engineering teams to plug these algorithms into closed-loop autonomous science pipelines.
• Set up benchmarks and ablation studies showing when and why evolved populations beat single-pass or non-iterative baselines.
• Shape the team's scientific direction through internal talks, reusable methods, datasets, prototypes and published papers.

Requirements

• A PhD in a quantitative field such as statistics, computer science, applied mathematics, computational biology or physics, with a strong research record. Equivalent expertise plus 5+ years of experience after a PhD is also acceptable.
• Deep experience in advanced Bayesian causal modeling, evolutionary programming, or both. Examples include probabilistic programming, Bayesian model comparison and weighting, genetic algorithms, quality-diversity methods and population-based search.
• A proven ability to design new algorithms from first principles and turn them into working, validated code.
• Strong software engineering skills and experience running computational experiments at scale, including parallel or distributed evaluation of many models.
• Recognized scientific leadership, such as first-author papers at top venues like NeurIPS, ICML or ICLR, influential publications, or widely used open-source tools.
• Experience applying these methods to biological, scientific or other mechanistic modeling problems is strongly preferred.
• Able to work as a senior IC: set technical direction, collaborate across research and engineering, and mentor junior colleagues.

Conditions

The salary range is $208,000 to $286,000. Final pay depends on qualifications, skills, competencies and experience.

Candidates who meet some but not all of the listed requirements are encouraged to apply.

The company is committed to an inclusive, diverse team.

Benefits

• Healthcare coverage, an annual incentive program, retirement benefits and a range of other benefits are offered. Benefit details are a good-faith estimate as of publication and may change.