Research Engineer - Geo-Distributed Inference
Pluralis Research2 months ago
Sydney, AustraliaSenior
Responsibilities
- Own the inference stack end to end, including pipeline-parallel execution, placement and routing, transport, serving-engine internals, and failure handling.
- Set the technical direction and build systems that remain fast and reliable as nodes join and leave, weights change, and hardware and networks are uncontrolled.
- Design, validate, and productionize new algorithms for inference over consumer hardware and the public internet.
- Maintain the rollout pipeline for reinforcement-learning training and evolve it into the serving layer for trained models.
Requirements
- Hands-on experience shipping serving-engine internals or building a large-scale inference system.
- Research ability demonstrated through publications, blog posts, or substantial unpublished work in distributed inference or related areas.
- Experience with low-bandwidth, high-latency systems operating over networks such as the public internet.
- Professional-level written and spoken English proficiency.
- Alignment with the goal of decentralized, trustless, and sovereign AI.
Benefits
- Equity-heavy package with significant ownership for key technical contributors and a high base salary.
- Remote-first work environment with globally distributed team members.
- Optional full visa sponsorship and relocation support to Australia or the United States.
- Remote work across Australia, North America, and other regions, with comfort working across time zones.
Categories
About Pluralis Research
Pluralis Research builds a protocol and tooling for collaborative, decentralized training and ownership of foundation models, enabling training and serving across heterogeneous, consumer-grade devices over the internet. The team demonstrated Agora, a permissionless run that pretrained an 8B model from scratch without any single participant holding the full weights. The privately held company is backed by Union Square Ventures and focuses on developer-facing infrastructure for large-scale model training and serving.