Research Engineer Intern

About the role

Pluralis Research is pioneering Protocol Learning, a decentralized approach to training large AI models on consumer‑grade devices. As a Research Engineer Intern you will join the Agora team, contributing to a production‑grade distributed training system that already powers an 8‑billion‑parameter model trained across the internet.

Key responsibilities

• Design, build, and ship a well‑scoped roadmap project on the Agora platform, culminating in a presentation to engineering and research teams.
• Develop and enhance multiprocessing, async I/O, and threading components within the distributed training stack.
• Work hands‑on with large‑scale training infrastructure on cloud providers such as AWS and GCP.
• Contribute daily to the production codebase, participating in code reviews and receiving mentorship from a dedicated buddy.

Required profile

• Current enrollment in or recent completion of a Master’s or PhD in machine learning, computer science, or a related field.
• Strong engineering ability with solid Python and PyTorch experience, especially in concurrent or parallel systems.
• Practical exposure to distributed machine learning and familiarity with cloud hyperscalers (AWS, GCP).
• Alignment with the vision of Protocol Learning as a trustworthy, decentralized AI pathway.

Required skills

• Python
• PyTorch
• Multiprocessing
• Async I/O
• Threading
• AWS
• GCP