P

Pluralis Research

Decentralized foundation-model training protocol

Research Scientist

AI ResearchConfirmed live in the last 24 hours
Full-time

About the Job

Requirements

  • A PhD in machine learning with publications in top-tier conferences such as NeurIPS, ICML, or ICLR is required.
  • Hands-on experience in large-scale distributed training and compression strategies.
  • Strong programming ability in PyTorch.
  • Professional-level written and spoken English proficiency.
  • Comfort working across time zones.

Responsibilities

  • Identify research questions that block Protocol Learning at scale, including communication efficiency, convergence under churn and staleness, heterogeneity, and robustness to malicious participants.
  • Solve open research problems and publish findings in Tier-1 venues.
  • Work with the engineering team to incorporate research results into live training runs.

Desired Qualifications

  • Experience with foundation model pre-training, post-training, or reinforcement learning.
  • Experience at proprietary, open-weight, or open-source AI labs.

About the Company

Pluralis Research

Decentralized foundation-model training protocol

Pluralis Research develops a protocol for decentralized training and shared ownership of foundation models. The core product is a framework that enables collaborative, distributed model development, allowing multiple parties to contribute data and compute without centralizing control. Revenue likely comes from licensing the protocol and offering consulting services to help organizations implement decentralized training methods. The team includes former FAANG researchers who bring deep expertise in AI research and scalable systems. Pluralis differentiates itself by focusing on decentralized, ownership-enabled training for foundation models rather than traditional centralized approaches, aiming to reduce reliance on centralized data processing. The company's goal is to make it practical for organizations to jointly train and own sophisticated foundation models through a distributed, cooperative framework.

Company size
11-50
Company stage
Seed
Total funding
$7.6M
Headquarters
San Francisco, CA
Founded
2024

Recently Posted Jobs

Explore more opportunities at Pluralis Research

See more jobs