Software Engineer Research Tools
Posted 2
views
Summary
Designs and builds full-stack internal tooling for Cursor's reinforcement-learning research: vendor task-submission and acceptance workflows, review tools for rollouts and grader outputs, and quality/versioning/catalog systems. Core stack is TypeScript and React with Node, Python, or Go.
You will design and build full-stack tools for researchers and external contributors to create, review, submit, and monitor reinforcement-learning environments and tasks. You will own review and acceptance workflows, build task-authoring interfaces, and create quality, versioning, and catalog tooling.
Responsibilities
- Create workflows for vendor task creation, submissions, and task acceptance into training
- Build review tools for rollouts, transcripts, grader outputs, and task-quality signals
- Develop environment-health, failure-search, versioning, and catalog experiences
- Establish a shared component kit and build self-service task-authoring interfaces with inline quality checks
Requirements
- Experience shipping full-stack products from user interface through storage or services
- Knowledge of TypeScript and React with Node, Python, or Go
- Experience building dense data-facing tools
- Experience building or maintaining a design system or component library
- Experience designing review, QA, moderation, fraud, or acceptance workflows
- Knowledge of data quality
