AfterQuery Training Data
Use expert-created examples and evaluation environments to train and test models.
AfterQuery Funding History
AfterQuery is an applied research lab that provides expert training data and post-training services for AI models. Its work includes datasets for supervised fine-tuning, reinforcement learning rubrics and environments for training agents.
Company Overview
AfterQuery
Founded 2024 · San Francisco, California, US
Official WebsiteAfterQuery Funding Timeline
Series A
US$30 million
AfterQuery framed the financing around capturing expert reasoning as training data for AI systems.
Led by Altos Ventures.
Source · afterquery.com ↗
AfterQuery Founding Team
Carlos Georgescu and Spencer Mateega met in high school at a Google summer program and later interned together at Meta. They started AfterQuery to build training data based on the work of specialists in fields where general-purpose datasets offered limited coverage.
AfterQuery Products
AfterQuery Company Development
Turning Legal Work into Repeatable Model Tests
AfterQuery’s collaboration with Legora took its evaluation work into legal practice. The teams used matters and supporting documents developed with participating firms to create tasks that could be run repeatedly against legal agents.
AfterQuery Competitive Landscape
Post-Training Supply Versus Inference Infrastructure
AfterQuery and Fireworks AI both serve teams working on model behavior, but at different stages. AfterQuery supplies expert examples, reinforcement-learning rubrics and agent environments. Fireworks gives teams infrastructure to serve models and evaluate or improve them alongside deployment.
