Research Engineer, Synthetic Data

About this role

Join a 15-person research engineering team building synthetic data infrastructure for AI agent training. You'll own the end-to-end pipeline that converts real-world workflows into high-quality, diverse training tasks using reinforcement learning and post-training methodologies.

What you'll do

• Build and maintain the synthetic data pipeline transforming domain workflows into training tasks for AI agents
• Collaborate with subject-matter experts to generate synthetic tasks across professional and technical domains
• Design task generation methods producing diverse, realistic, and learnable outputs at scale
• Create tooling to validate, mutate, and iteratively improve synthetic tasks
• Analyze agent performance on synthetic tasks to identify learning gaps and failure modes
• Develop metrics quantifying task diversity, realism, learnability, and quality

What they're looking for

• Python programming
• Data pipeline design and implementation
• Synthetic data generation and evaluation
• Docker and Linux containerization
• ML/AI infrastructure development
• Evaluation frameworks and benchmarking
• Reinforcement learning paradigms
• Large-scale dataset processing and validation

Benefits

• Visa sponsorship available
• Equity participation in early-stage startup
• High ownership with minimal bureaucracy
• Work with published researchers and Olympiad medalists
• Unstructured problem-solving environment