Description
Pluralis Research is hiring an Inference Systems Engineer to build and own the end-to-end inference stack for its decentralized Protocol Learning system. The role focuses on pipeline-parallel execution, placement and routing, transport, serving engines, failure handling, and the development of novel algorithms for fast inference on consumer hardware over the public internet. The engineer will support reinforcement-learning rollouts today and prepare the system for model serving later, with a remote-first culture, global team distribution, and optional visa sponsorship to Australia or the United States.
