Posted a month ago
Description
Pluralis Research is seeking a seasoned RL post-training engineer to build the end-to-end post-training stack, from rollout ingestion to policy updates and weight synchronization across a geo-distributed network. You’ll shape algorithms and systems to make decentralized, asynchronous post-training work at scale.
We value hands-on RL post-training, strong Python/PyTorch engineering, and a research mindset. This is a remote-first role with global collaboration across time zones.
#J-18808-Ljbffr
