Research Engineer, Synthetic Data

The role

You will build pipelines that turn professional and technical workflows into realistic training tasks for AI agents. You will work with subject-matter experts across domains and analyze how agents perform on these tasks.

What you will do

• Build pipelines that convert professional and technical workflows into realistic, structured training tasks.
• Collaborate with subject-matter experts to develop tasks across distinct domains.
• Design methods for generating diverse, realistic, and learnable synthetic tasks.
• Create tools to mutate, validate, and improve generated tasks.
• Analyze agent performance to understand task effectiveness and identify failure modes.
• Develop metrics for task diversity, realism, learnability, and overall quality.

Required qualifications

• Two or more years of experience in software engineering, machine learning engineering, or AI research, building and delivering data pipelines, ML infrastructure, or synthetic data systems.
• Technical education or equivalent technical background, with strong Python skills and experience in Linux environments and Docker.
• Hands-on experience applying synthetic data methods to build end-to-end generation pipelines for AI or machine learning.
• Understanding of synthetic data quality, evaluation criteria, and limitations.
• Experience building evaluation frameworks, benchmarks, or testing environments for AI agents or large language models.
• Ability to independently deliver projects in open-ended settings, with careful attention to data quality and edge cases.

Nice to have

• Familiarity with reinforcement learning, agent workflows, or LLM post-training.

Before you apply

• Visa sponsorship is available.

About the company

clera is building a small engineering team focused on synthetic data systems for AI agents. The team works to improve how models learn and perform in real-world applications.