Applied Scientist

Research Scientist

Building the learning systems behind superintelligence.

Location

: Hybrid / Remote

(San Francisco, CA \& Cambridge, MA Sites)

Seniority

: L4 - \> L8

Compensation

: Highly Competitive (DOE) – Base, Bonus \+ Equity

Overview

A Pioneering AI Lab building self-learning AI systems to solve humankind's greatest challenges.

The world's best AI systems can write code, solve complex mathematics, reason through scientific papers, and outperform experts on a growing list of benchmarks. Yet they still can't outperform the average graduate at generating a genuinely useful scientific discovery. This isn't an intelligence issue, but a lack of physical laboratory and ability to close the thought - \> experiment loop.

Our team is building a platform that combines frontier AI models, autonomous experimentation, robotics, and scientific infrastructure into a single learning system capable of executing the scientific method at unprecedented scale.

The Opportunity

The Frontier Capabilities Research Team is a deep, talent dense ecosystem working across reasoning, post-training, reinforcement learning, agents, and open-ended learning. These roles span a few complementary directions.

Candidates are expected to bring deep expertise in one (or more) of the following areas..

1. Agentic System Building:

Build systems that autonomously propose, execute, and verify scientific hypotheses over long time horizons

1. Distillation

: Translate strong inference-time behaviors and reasoning traces into efficient, trainable models

1. Scalable Experience Generation

: Develop inference-time algorithms and synthetic data pipelines that generate high-quality training signal for scientific reasoning

1. Model Experimentation

: Own, experiment with, and train large foundation models

1. Evaluations \& Benchmarks:

Designing and building SOTA scientific agentic benchmarks and harnesses whilst owning post-training end-to-end

About You:

• MS or PhD in CS, ML, AI, or Equivalent
• Strong foundations with LLMs and proven experience training models with at least 30B\+ parameters
• Experience designing / executing ML experiments, including benchmarking, ablations, \& large-scale training / evaluation pipelines
• Ability and interest in defining and pursuing research directions in open-ended, rapidly evolving spaces

You’d Stand Out With:

• Experience with synthetic data generation, distillation, or self-improvement loops
• Familiarity with reinforcement learning
• Experience with planning, search, or decision-making systems at scale
• Experience in building agentic systems or multi-agent workflows

For more information or an informal discussion apply today!