Strong familiarity with LLM training and evaluation methodologies.
Genuine obsession with data structure, selection, and quality in relation to model behavior.
Ability to design lightweight experiments quickly and derive actionable insights from messy results.
Comfortable working across diverse domains including finance, software engineering, and policy.
Proactive mindset with a preference for building rather than theorizing.
Undergraduate or master’s research background preferred, PhD not required.
Responsibilities
Run controlled SFT and RL experiments to evaluate dataset impact on models.
Quantify performance improvements across various capabilities.
Communicate findings to partner labs to support sales efforts.
Collaborate with internal teams to enhance data quality based on experimental results.
Benefits
Collaborative work environment with cutting-edge technology.
Access to diverse projects spanning multiple domains.
Opportunities for personal and professional growth through experimentation.
Full Job Description
The Role
Your job is to prove that our data works. You will design and run training experiments that isolate the impact of our datasets on model behavior. This includes SFT and RL-based post-training, where you'll measure how different data sources shift capability, generalization, and alignment. Working closely with partner labs, you will turn our datasets into clear, defensible evidence: this data this improvement under these conditions. This is experimental, high-leverage work.
What You'll Do
Run controlled SFT and RL experiments to measure the impact of our datasets on model performance.
Help build public evals and new data types that push the frontier.
Publish external-facing research, blog posts, and technical reports.
Work with internal SPLs to iterate on data quality based on your results.
What We're Looking For
Strong familiarity with LLM training and evaluation methodologies.
Genuine obsession with how data structure, selection, and quality drive model behavior.
Ability to design lightweight experiments, move fast, and extract actionable insights from messy results.
Comfort working across domains (you'll touch finance, software engineering, policy, and more).
A bias toward building over theorizing.
Great candidates are undergrad research or master's research (but haven't done a phd).
