Research Scientist - Post Training

Afterquery

After Query is an applied research lab curating data solutions for foundation model development.

We serve every frontier AI lab with the mission of delivering the best data to power the best models. In doing so, we can make expertise that once took a lifetime to build available to anyone who needs it. Our customers are the ones building the foundation models themselves and our work sits directly in the loop of how those systems improve.

This is a rare opportunity to join a company at a defining moment in AI. Since raising our $30M Series A at a $300M valuation, After Query has grown well over a $100M revenue run rate.

We're based in San Francisco and backed by leading investors including Altos Ventures, Box Group, and Y Combinator and angels from Google Deep Mind, OpenAI, Anthropic, Meta Superintelligence Labs, and Microsoft AI.

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.

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.

Strong familiarity with LLM training and evaluation methodologies.

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).

Genuine obsession with how data structure, selection, and quality drive model behavior.