Research Scientist - Post Training

About Us

We build training data and evaluation infrastructure that frontier AI labs use to improve their models. We partner with the world's leading labs to design high-signal datasets and run rigorous evaluations that go beyond static benchmarks. We're a small, early team (post–Series A) where individual contributors have direct impact on how the next generation of models learns and improves.

The Role

We're building out our post-training research team and hiring 2–3 Research Scientists to work together on this mission. Your job is to prove that our data works. You'll design and run training experiments that isolate the impact of our datasets on model behavior, including SFT and RL-based post-training, to measure how different data sources shift capability, generalization, and alignment. Working closely with partner labs, you'll turn our datasets into clear, defensible evidence: this data this improvement under these conditions. It's experimental, high- leverage work at the edge of model development.

What You'll Do

1. Run controlled SFT and RL experiments to measure the impact of our datasets on model performance.
2. Quantify lift across capabilities — reasoning, tool use, long-horizon tasks, and domain-specific workflows. Share findings directly with partner labs to deepen relationships and drive sales.
3. Collaborate with internal SPLs to iterate on data quality based on your results.
4. Work closely with the other Research Scientists on this team to build shared experimental infrastructure and benchmarks.

What We're Looking For

1. Strong familiarity with LLM training and evaluation methodologies (SFT, RL post-training).
2. Genuine obsession with how data structure, selection, and quality drive model behavior.
3. Ability to design lightweight experiments, move fast, and extract actionable insights from messy results. 
4. Comfort working across domains — you'll touch finance, software engineering, policy, and more.
5. A bias toward building over theorizing.

Must-Have Requirements

1. Strong familiarity with LLM training and evaluation methodologies, including SFT and RL post-training.
2. Genuine obsession with how data structure, selection, and quality drive model behavior.
3. Ability to design lightweight experiments, move fast, and extract actionable insights from messy results. Comfort working across domains — finance, software engineering, policy, and more.
4. Undergrad or master's research background; pre-PhD candidates preferred.

Nice-to-Have Requirements

1. Prior work or internship at an RL environment company, AI safety org, or benchmarking org (METR, Artificial Analysis, or equivalent).
2. Experience running controlled training experiments end-to-end.
3. Published research on model evaluation, post-training, or data curation.
4. Strong SWE chops alongside research instincts. Compensation

Compensation

$250K–$450K total compensation equity 

Requirements

1. Run controlled SFT and RL experiments to measure dataset impact on model performance
2. Quantify lift across capabilities including reasoning, tool use, long-horizon tasks, and domain- specific workflows
3. Communicate findings with partner labs to drive sales
4. Work with internal SPLs to iterate on data quality based on experimental results
5. Strong familiarity with LLM training and evaluation methodologies
6. Design lightweight experiments and extract actionable insights from messy results 
7. Work across multiple domains including finance, software engineering, and policy