Mirendil
Member of Technical Staff, Post-Training, RL
$300K–$400K
San Francisco · On-site · Staff · Full-time · Posted Jun 24, 2026
About the job
Build and run experiments for the post-training stack that makes frontier reasoning models solve difficult tasks reliably and at scale.
What you'll do
- Develop and iterate on RL, SFT, and distillation recipes, including objectives, data mixtures, and rollout strategies.
- Scale RL training and develop methods for long-horizon reasoning, off-policy training, and robust verification.
- Analyze and debug experiments, partnering with systems, infrastructure, and data teams to make successful runs reliable and repeatable.
Pay
$300,000–$400,000 USD (base salary)
Tools
RL, SFT
AI focus
Train frontier reasoning models through post-training, including reinforcement learning, supervised fine-tuning, distillation, verification, and reward-quality improvements.
Model-written summary · the original posting is authoritative