Software Engineer, RL Data

Cursor
San Francisco
Workplace: OnsiteFull timeFunction: Software EngineeringSkills: ["Truth-seeking","Passionate","Creative","Spirited debate","Iterative problem-solving","Turning messy inputs into datasets/tools"]

Create the tasks, rewards, and environments that train Cursor’s coding agents on real user data. Own how RL data is defined and scored, then iterate from traces and evals until models improve. Investigate failure modes, build systems that surface key behaviors, and turn one-off setups into reusable components for other teams. Partner with research to verify datasets teach the intended capabilities.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
Cursor
Cursor
2 days ago

Software Engineer, RL Data

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 10 hours agoStatus: Live

Job Summary

Create the tasks, rewards, and environments that train Cursor’s coding agents on real user data. Own how RL data is defined and scored, then iterate from traces and evals until models improve. Investigate failure modes, build systems that surface key behaviors, and turn one-off setups into reusable components for other teams. Partner with research to verify datasets teach the intended capabilities.
Location: San Francisco
Workplace: Onsite
Employment Type: Full time
Job Function: Software Engineering

Key Responsibilities

  • •Design task sets that teach specific agent capabilities, iterating from traces and evals until performance improves.
  • •Analyze agent traces to identify failure modes or surprising behaviors, and build systems that surface similar cases.
  • •Transform one-off recipes into reusable solutions (rewards, environments, and improved data quality).
  • •Partner with research to test whether a dataset is teaching the intended capability.

Key Requirements

  • •Write careful, fast code with strong software engineering fundamentals.
  • •Like setting tasks by breaking fuzzy capabilities into measurable objectives.
  • •Have an infra, data, or distributed systems background; RL experience is a plus.
  • •Enjoy analyzing messy real-world agent behavior and converting it into a dataset or tool.
Experience:Reinforcement learningMachine learningDistributed systems
Skills:Truth-seekingPassionateCreativeSpirited debateIterative problem-solvingTurning messy inputs into datasets/tools
Tech Stack:Reinforcement learning (RL)TasksRewardsEnvironmentsAgent tracesEvalsDatasetsDistributed systemsData qualityData scoringFailure mode analysis

Company Brief

Cursor
Builds Cursor, an AI-powered development platform and code editor that helps developers write, review, and maintain code using large language models and developer-focused automation tools for enterprise engineering teams.
Industry: Developer Tools
Company Size: Large (251 to 1,000 employees)
Revenue: USD 1B+
Growth: Scaleup
Valuation: Unicorn (USD 1B+)
Funding: Series D
Headquarters: San Francisco, United States
Founded: 2022
Glassdoor
Glassdoor: 3.0
WebsiteLinkedInGlassdoor