What to know
About AfterQuery Experts
AfterQuery is a San Francisco applied-research company founded in 2025 by Carlos Georgescu and Spencer Mateega. It develops evaluations and training environments for computer-using AI agents, while AfterQuery Experts is the contributor network that supplies professional judgment, reference work and model feedback.
AfterQuery also builds AI products and research tools, but this profile focuses on the experience of people applying through the Experts network.
AI data training work at AfterQuery Experts
AfterQuery’s distinguishing work is the creation and review of long-horizon trajectories: realistic sequences showing how a professional or software agent should complete a task, recover from errors and verify the result. Its public research includes Terminal-Bench and other agent evaluations. Contributor projects can therefore involve inspecting repositories, testing code, documenting tool use, writing reference solutions, evaluating research and building scoring rubrics, alongside more conventional annotation and response comparison.
Some opportunities are worldwide; others restrict countries or professional backgrounds. Rates can be hourly or task-based and may differ substantially by specialty, so only rely on the figure attached to the role you choose.
What is it like to work at AfterQuery Experts?
Work is remote and project-based. Some assignments allow flexible scheduling, while others request a weekly commitment or delivery window. You are not guaranteed ongoing tasks simply because your expert profile is approved. AfterQuery can keep qualified contributors available until a client project opens, and demand can pause between engagements.
The work is best suited to people who can operate independently, document their reasoning and follow detailed evaluation criteria. Some project descriptions remain broad because client material and model behavior are confidential; the project offer should still identify the deliverable, time commitment, rate and approval rules before production begins.
Recent developments and research
In April 2026, AfterQuery announced a $30 million Series A at a $300 million valuation and said it had surpassed a $100 million revenue run rate. In July, the company published details of how NVIDIA used AfterQuery’s Office Agent trajectories to improve Nemotron 3 Ultra on GDPval-style professional tasks. Together, these developments reinforce that the network’s long-horizon, file-based expert work is being used in frontier model training and evaluation.
Application process at AfterQuery Experts
The application process begins with an account or a role in the application directory. The process can ask for a CV, work history and questions about your specialty. AfterQuery describes screening that may include an asynchronous video interview or a skills assessment.
Passing makes you eligible for project matching and does not always mean work starts immediately. Selected contributors complete contractor, confidentiality, identity and payment setup, followed by any project calibration. The project-specific offer confirms the rate, acceptance criteria and payout method before billable production begins.
Sources: Company website · Terms · Privacy · Series A and company milestone · NVIDIA GDPval research
Information last updated August 2026.