← Search all jobs

Transparent Search Group Member of Technical Staff, Head of Quality San Francisco, California, United States $180K–$280K View role

Transparent Search Group

San Francisco, CA · Posted September 23, 2026

See every open role at Transparent Search Group

More in California job market · Engineering jobs

About the job

Company: Confidential - Venture-backed AI research lab

Location: San Francisco, CA (SoMa office, 5 days per week)

Compensation: $180,000 - $280,000 + competitive equity

Employment Type: Full-time

Visa Sponsorship: Visa transfers (OPT, H-1B transfer)

About the Company

An applied research lab building infrastructure that turns real-world data into simulated environments for training specialized AI agents, supporting frontier labs and enterprises.

The Role

Our client is hiring a Head of Quality (1-5 years) to own the quality bar for every environment and dataset it delivers, and to build the systems and team that enforce it at scale. This is hands-on: you will be deep in the data reviewing samples, building automated QA, and making sure what ships produces real learning signal. You will also hire and lead the QA team as the company scales.

What You Will Do

Review human data and environment outputs hands-on to keep quality extremely high.

Build automated evaluation, QA agents and grading systems that catch reward hacking, broken verifiers and distribution mismatches before delivery.

Define quality metrics and acceptance criteria for environments, tasks, verifiers, rewards and datasets.

Own delivery acceptance for engagements with frontier labs, hyperscalers and enterprises.

Hire, onboard and manage QA engineers and reviewers.

What You Bring

1-5 years of experience in ML, RL, data quality or evaluation

Understanding of what a dataset actually teaches a model, beyond checklists

Experience building automated QA or evaluation systems

Ability to work in the San Francisco office 5 days a week

Tech Stack

Machine learning, reinforcement learning, LLMs, data quality, QA automation, evaluation frameworks