1 day ago
Base Salary
$180k - $500k/yr
Responsibilities
- Work on post-training and RLVR pipelines to assess how datasets, rewards, and training strategies affect model performance.
- Design reward-shaping experiments and algorithmic improvements such as GRPO and DAPO for LLM tool use, agentic behavior, and reasoning.
- Measure data usability, quality, and performance improvements against benchmarks.
- Build and maintain scalable data generation and augmentation pipelines.
- Create rubrics, evaluators, scoring frameworks, LLM evaluation systems, benchmarks, and metrics.
- Collaborate with AI researchers, applied AI teams, and training-data experts in a rapid experimental environment.
Requirements
- Strong applied research background focused on post-training and/or model evaluation.
- Strong coding proficiency and hands-on experience with machine learning models.
- Strong understanding of data structures, algorithms, backend systems, and core engineering fundamentals.
- Familiarity with APIs, SQL/NoSQL databases, and cloud platforms.
- Ability to reason deeply about model behavior, experimental results, and data quality.
- Industry post-training experience, model training or evaluation experience, synthetic data generation, LLM evaluation, RL-style workflow experience, top-tier conference publications, and relevant work samples are preferred.
Benefits
- $1.5K monthly meal stipend.
- Free Equinox membership.
- $200 monthly laundry reimbursement.
- $200 monthly personal wellness reimbursement.
- Work in person five days per week in San Francisco, NYC, or London, with optional remote Saturdays.
Tech Stack
Categories
AI ResearchML Engineering
About Mercor
Mercor builds an expert-powered platform that trains, evaluates, and deploys AI systems for AI labs and enterprises. It operates APEX to assess model performance on economically valuable tasks, and provides services such as custom AI agents and teams of vetted domain experts who encode organizational knowledge into AI. Founded in 2023 and headquartered in San Francisco, it is privately held and works with frontier AI labs and enterprise clients requiring strict data isolation.