July 6, 202610 min read

Mercor Founders

Woman at desk arranging three connected prototype objects in sequence, tracing flow between them with finger as dusk light st

Mercor: An AI Evaluation Platform

Mercor is an AI evaluation platform that connects qualified contractors with AI companies needing human feedback on model outputs. The platform specializes in domain-specific evaluation services including mathematics, coding, scientific reasoning, and creative writing, distinguishing itself through contractor quality screening rather than high-volume crowdsourced annotation.

Company Overview

Mercor operates as a two-sided marketplace connecting specialized contractors with AI evaluation projects. The platform screens contractors through domain-specific assessments to ensure quality evaluation work. As of publicly available information, the company's contractor network supports AI training and evaluation services for multiple AI companies.

About the Founders

Mercor was founded by Brendan Foody (CEO), Adarsh Hiremath (COO), and Surya Midha (CTO). The founders met through the debate program at Bellarmine College Preparatory, a Jesuit high school in San Jose, California.

Their competitive debate experience directly translates to AI evaluation work. Policy debate trains participants in rapid research, argument evaluation, evidence quality assessment, and decision-making under time pressure. These skills apply directly to evaluating AI model responses where contractors assess factual accuracy, reasoning quality, and instruction-following capabilities.

Company Funding and Valuation

Mercor has raised capital across Seed, Series A, and Series B funding rounds, with a significant Series B milestone in April 2023.

Platform Operations and Differentiation

Mercor builds direct relationships with AI companies rather than operating as a middleman. The platform emphasizes contractor quality and domain expertise screening over crowd-sourced volume. Its focus remains exclusively on AI training data and evaluation services rather than broader AI infrastructure.

Actionable Insights for Aspiring AI Evaluators

If you are considering a career evaluating AI models, take these concrete steps:

Develop expertise in specific domains (mathematics, coding, scientific reasoning, or creative writing) that AI companies need evaluated.
Learn RLHF fundamentals, rubric engineering, response quality assessment, and fact-checking techniques, as these skills directly match what platforms like Mercor screen for.
Obtain formal credentials through the AI Evaluator Certification from Annotation Academy (24 modules, 50+ hours, 800+ practice questions) to qualify for contractor networks requiring domain-specific assessment.
Build a track record demonstrating quality evaluation judgment and reasoning assessment capabilities, as platforms prioritize contractor quality over volume.
Research specialized evaluation platforms that focus on your domain expertise, as compensation and opportunity vary based on project type and required skills.

Compensation for evaluation work varies based on project type, domain expertise, and platform. Formal certification and demonstrated domain knowledge increase your qualification level for higher-value evaluation projects.