Researcher, Computer Use - Agent Post-Training

Train and evaluate agents to reliably operate computers and improve model behavior.

• •Work on Agent Post-Training to teach models to operate computers, improve agentic behavior, and ship capabilities into products.
• •Key Responsibilities Design and run experiments to improve computer-use agent behavior.
• •Own post-training stack: RL, data pipelines, graders, reward signals, and evaluations.
• •Build evals/environments and convert failures into training data or product fixes.
• •Partner with product and infrastructure teams to translate signals into model improvements.
• •Requirements Strong technical fundamentals in ML, software engineering, systems, or statistics.
• •Hands-on experience with LLMs, RL, RLHF/RLAIF, post-training and evals.
• •Ability to design experiments, build pipelines, and analyze model behavior.
• •Comfort working cross-functionally with research, product, infra, and safety teams.