Research Engineer, Synthetic Data
Core
Build pipelines to convert professional workflows into realistic, structured training tasks for AI agents.
Role type
Research Engineer, Synthetic Data
Builds
Scalable training tasks and evaluation frameworks for AI agents
Domain
AI/ML, Synthetic Data, Agent Workflows
Deliverable
production ML models
Required skills
Python, Linux, Docker, Synthetic Data Generation, Pipeline Development, Evaluation Frameworks, Data Quality Analysis
Preferred skills
Reinforcement Learning, LLM Post-Training, Agent Workflows
Technologies
Python, Linux, Docker
Responsibilities
Build pipelines converting workflows to training tasks, Design methods for generating diverse synthetic tasks, Create tools to mutate and validate generated tasks, Analyze agent performance to identify failure modes, Develop metrics for task quality and learnability, Collaborate with subject-matter experts on domain-specific tasks
Seniority
Mid-level, hands-on IC