Head of RL Gym - Program Operations

Location: Ramat Gan, IL
Salary range: Not stated
Workplace: In office
Level: Manager
Job family: T&S Operations
Posted: Recently

The T&S Network estimates this role to be M4, Senior Manager. We don't yet know typical pay for this level in Israel. Most of our pay data comes from US roles. If you work in Trust & Safety in Israel, add your pay to help others. Key skills at this level: forecasts volume and staffs against it with real error bars, owns service levels and can explain a miss without blaming the queue, and governs a vendor relationship on quality, not only on cost per action.

Learn more about this and similar levels

Why this level

Estimated from the posting rather than stated by the employer: the title reads as M4 · asks for 7+ years.

Apply on Alice

The role, in T&S terms

Written by the T&S Network from Alice's posting.

Owns end-to-end delivery, quality, and scaling of RL Gym programs built for strategic AI partners, with frontier-model teams. The role sits across Delivery, Product, Sales, Research, and Operations to turn customer and model-development needs into scalable programs and to build the program management and operations function as the business grows.

What you'd do

Run partner programs from scoping through launch, day-to-day execution, scaling, and iteration.
Turn research and model needs into program requirements, workflows, and success metrics.
Work with frontier-model teams to find model gaps and design RL Gym programs to close them.
Set and enforce quality standards and quality assurance across programs.
Partner with Sales on discovery, solution design, scoping, and expansion.
Pass partner needs and operational learnings to Product as requirements and roadmap priorities.

What it asks for

Background running complex, cross-functional programs at scale.
Proven record building scalable operating processes while holding a high quality bar.

Terms in this role

Red teaming.: Testing a system, often an AI model, by deliberately trying to make it fail or misbehave, so problems are found and fixed first.
RLHF.: Reinforcement learning from human feedback: training an AI model on people's ratings of its answers.

Our best summary of the posting, and it can be wrong. The employer's posting is what counts: read it in full on Alice before you apply. If we got something wrong, report this listing.

Tired of job hunting?

Let the roles come to you. Switch on Available, and when a hiring manager or recruiter brings us a role that fits, we'll put you forward. Never to the company you work for.

Join Hidden Talent

Still listed 14 hours ago