Engineering

Marketing

Business Development

Administrative

Research

Entrepreneurship

Decentralized collaborative training protocol for foundation models

Founder

2 Aussie startups make the cut for the latest AWS Generative AI accelerator

AWS taps two Australian startups for its 2025 Generative AI Accelerator

Research Scientist - Decentralized Training and Inference Verification

Pluralis Research develops decentralized training and inference for large AI models across consumer-grade devices. The Research Scientist will design efficient algorithms and systems to verify model inference and training contributions in a trustless network.

Research Scientist - Post-Training

Pluralis Research is developing decentralized training and serving for large models across consumer devices. The Research Scientist will build the reinforcement-learning post-training stack and algorithms for asynchronous, high-latency decentralized infrastructure, and deliver the first post-trained models.

Research Scientist - Geo-Distributed Inference

Pluralis Research develops decentralized systems for training and serving large models across consumer devices. The Research Scientist will build and own a fast, reliable geo-distributed inference pipeline, invent and productionize inference algorithms, and support reinforcement-learning training rollouts and future model serving.

Research Scientist - Pre-training

Pluralis Research develops decentralized training and serving of large AI models across consumer-grade devices. The Research Scientist will build and optimize distributed pretraining systems for frontier-scale models, including communication efficiency, fault tolerance, and run monitoring across heterogeneous hardware and networks.

Machine Learning Engineer - ML Training Platform

Pluralis Research develops Protocol Learning for decentralized training and serving of large models. The Machine Learning Engineer will architect, build, and scale infrastructure orchestration, distributed compute, and supporting services for continuous experimentation and large-scale training across consumer nodes and cloud instances.

Research Engineer Intern

Pluralis Research develops Protocol Learning, a decentralized approach to training and serving large models on consumer-grade devices. The Research Engineer Intern will work on Agora, the company’s production decentralized training system, shipping roadmap projects and contributing to concurrent systems, distributed training infrastructure, and the production codebase.

Research Engineer - Geo-Distributed Inference

Pluralis Research is hiring a Research Engineer to build and own an inference pipeline for decentralized AI models running on consumer hardware over the public internet. The role focuses on developing fast, reliable inference systems and algorithms for reinforcement learning training rollouts and future model serving.

Research Engineer - Decentralized Training and Inference Verification

Pluralis Research develops decentralized systems for training and serving large models across consumer-grade devices. The Research Engineer will develop efficient algorithms and systems to verify inference outputs and training contributions in its trustless network.

Research Engineer - Pre-training

Pluralis Research develops Protocol Learning systems for decentralized large-model training and serving. The Research Engineer will build and scale distributed pretraining systems for large models across heterogeneous hardware and internet-connected regions, focusing on performance, fault tolerance, and monitoring.

Research Engineer - Post-Training

Pluralis Research develops decentralized training and serving for large models. The Research Engineer - Post-Training will build the end-to-end reinforcement-learning post-training stack, develop algorithms for asynchronous and high-latency training, and release the first decentralized post-trained models.