Research Engineer Intern
Pluralis Research2 months ago
Remote, AustraliaIntern
Responsibilities
- Design, build, and ship a well-scoped project on the Agora roadmap and present the results to engineering and research teams.
- Build and improve multiprocessing, asynchronous I/O, and threading components in a production distributed training stack.
- Work hands-on with large-scale training infrastructure across AWS and GCP.
- Contribute daily to the production codebase with code review and mentorship from an Agora team member.
Requirements
- Be currently enrolled in or have recently completed a Master’s or PhD in machine learning, computer science, or a related field.
- Have strong Python and PyTorch skills.
- Have experience building concurrent or parallel systems using multiprocessing, asynchronous I/O, or threading.
- Have hands-on distributed machine learning exposure through an internship, coursework, or serious projects.
- Have experience with AWS, GCP, or other hyperscalers.
- Have professional-level written and spoken English proficiency.
- Top-tier machine learning publications are welcome but not required.
- Open-source contributions to machine learning frameworks, distributed systems, or networking libraries are a plus.
Benefits
- Full-time, fixed-term three-month internship with an option to extend.
- Remote work across time zones with teams in Australia and North America.
- Australia-based candidates only.
- Work on live production systems with mentorship and ownership of a genuine open problem.
Tech Stack
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.