Design synthetic company environments with realistic email, records, and files for extended agent use.
Create graders and evaluation systems to assess AI agents.
Develop benchmarks that correlate with real-world production performance.
Find and fix evaluation failures, including reward hacking and benchmark gaming.
Summarised from the posting by AI and can be wrong. The full description is the source.
Requirements
Experience
Hands-on experience with synthetic data pipelines, benchmarks, evaluations, or post-training
Experience building environments or benchmarks used to train or evaluate agents, and designing graders or evaluation systems for AI agents
Experience simulating complex software systems for agent tool use, and identifying benchmark flaws or fixing reward hacking
At least 1 to 2 years at a fast-growing or AI-focused startup, with relevant AI research or engineering experience
Strong startup experience, such as founding a company or participating in an accelerator
High-level competition experience in areas such as mathematics, programming, debate, or games
Other requirements
Conviction that AI agents can outperform humans at many computer-based tasks
Benefits
Equity
Meaningful founding equity
Bonus
Commission-style bonus of up to $200,000 or more, tied to the value of delivered environments
Full job description
From Clera’s careers site
ABOUT THE ROLE
Build evaluation environments and benchmarks that measure whether AI agents can perform real company work and match or exceed human performance. This founding engineering role focuses on agent capability assessment and production readiness, shaping how agent competence is measured.
WHAT YOU'LL DO
Design synthetic company environments with realistic email, records, and files for extended agent use.
Create robust graders and evaluation systems that can assess agents beyond the expertise of human reviewers.
Develop benchmarks that correlate with real-world production performance.
Find and fix evaluation failures, including reward hacking and benchmark gaming.
Show full descriptionShow less
Simulate software tools such as inboxes and CRMs with enough fidelity to support agent interactions at scale.
WHAT WE'RE LOOKING FOR
Hands-on experience with synthetic data pipelines, benchmarks, evaluations, or post-training; this is a research and engineering role, not a pure infrastructure position.
Experience building environments or benchmarks used to train or evaluate agents, and designing graders or evaluation systems for AI agents.
Experience simulating complex software systems for agent tool use, and identifying benchmark flaws or fixing reward hacking.
At least 1 to 2 years at a fast-growing or AI-focused startup, with relevant AI research or engineering experience.
Strong startup experience, such as founding a company or participating in an accelerator, and high-level competition experience in areas such as mathematics, programming, debate, or games.
Conviction that AI agents can outperform humans at many computer-based tasks.
COMPENSATION & BENEFITS
$200,000 to $300,000 annual base salary, with up to $350,000 for exceptional candidates. An additional commission-style bonus of up to $200,000 or more is tied to the value of delivered environments. Meaningful founding equity is available. The role description states that visa sponsorship is available.
LOCATION
On-site five days a week in San Francisco, California, United States.