Research Scientist - Post-Training

The role

You will build the RL post-training stack and invent algorithms to make RL work on consumer GPUs and Macs spread across the public internet. You will also build evaluations and ship the first decentralized post-trained models.

What you will do

• Build the RL training loop end-to-end including rollout ingestion, reward computation, policy updates, and weight distribution
• Adapt RL recipes to asynchronous, high-latency, partially trusted generation
• Build evaluations to show model improvement
• Take the first decentralized post-trained release from run to public artifact

Required qualifications

• Hands-on RL post-training on large language models such as RLHF, RLVR, or reasoning-focused RL
• Experience with the systems layer including rollout generation, async training loops, and weight synchronization
• Production-quality Python and PyTorch
• Professional-level English proficiency (written and spoken)
• Comfortable working across timezones

Nice to have

• Experience training over slow networks, or with decentralized or federated setups
• Familiarity with serving-engine internals such as vLLM or SGLang
• Experience with reward modeling or building verifiable-reward datasets
• Experience with P2P networking and NAT traversal
• Experience at proprietary, open-weight and open-source AI labs
• Publications in RL post-training, asynchronous or distributed RL, or nearby fields

Benefits

• Equity-Heavy Package
• High base salary
• Remote-First Culture
• Optional full visa sponsorship
• Relocation support to either Australia or the US

About the company

Pluralis Research works on Protocol Learning to train and serve large models in a fully decentralized way on small consumer-grade devices. They are a world-class, deeply technical team of ML researchers backed by Union Square Ventures and other tier-1 investors.