Research Engineer, Intern – Benchmarking

Mercor•San Francisco, CA

9d•$180,000 - $500,000•Onsite

About The Position

Mercor is hiring a Research Engineer Intern to work at the intersection of engineering and applied AI research. The role owns benchmarking pipelines, evaluation systems, and failure-analysis workflows that inform how frontier language models are trained and improved. The work defines how Mercor measures tool use, agentic behavior, and real-world reasoning, and involves designing and running evaluations, building rubrics and scorers, and translating failure analysis into improvements for post-training, RLVR, and data pipelines. The role is based in person five days a week. The posting describes Mercor as a profitable Series C company valued at $10 billion and says its offices are in San Francisco, New York City, and London. The role's stated candidate preference is to work in person in San Francisco five days a week.

Requirements

Strong applied research background focused on model evaluation, benchmarking, and/or failure analysis.
Strong coding skills and hands-on experience with ML models and evaluation code.
Solid grasp of data structures, algorithms, and backend systems.
Comfort with APIs, SQL/NoSQL, and cloud platforms for running and storing evaluation results.
Ability to reason about model behavior, experimental results, and data quality from evaluations and failure analyses.
Excitement to work in person in San Francisco five days a week in a high-intensity, high-ownership environment.

Nice To Haves

Industry experience on a post-training or evaluation/benchmarking team (highest priority).
Publications at top-tier venues such as NeurIPS, ICML, or ACL, especially in evaluation or benchmarking.
Experience building or running LLM evaluations, benchmarks, or failure-analysis pipelines.
Experience with synthetic data generation, rubric design, or RL-style workflows that use evaluations for reward shaping.
Work samples or code demonstrating relevant skills, such as evaluation frameworks, benchmark suites, failure-analysis reports, or tooling.

Responsibilities

Design, implement, and maintain benchmarks and metrics for tool use, agentic behavior, and real-world reasoning; ensure benchmarks scale with training and stay aligned with product and research goals.
Build and operate end-to-end LLM evaluation systems, including runs, scoring, dashboards, and reporting, so teams can track model performance and compare runs at scale.
Conduct systematic failure analysis of model outputs, categorize failure modes, quantify their prevalence, and feed findings into reward design, data curation, and benchmark design.
Create and refine rubrics, automated evaluators, and scoring frameworks for training and evaluation decisions, balancing rigor with scalability and considering human versus model-as-judge approaches, calibration, and agreement.
Quantify data usability, quality, and impact on key benchmarks; use evaluations and failure analysis to guide data generation, augmentation, and curation.
Collaborate with AI researchers, applied AI teams, and data producers to align evaluations with training objectives and prioritize important benchmarks and failure analyses.
Take strong ownership of benchmarks, evaluations, and failure-analysis workflows in a fast-iteration research environment.