Pluralis Research
Research Engineer - Geo-Distributed Inference
- Confirmed live in the last 24 hours
- No salary listed
- Mid level
- Full-time
- Remote · San Francisco
- Added 1 month ago
About this role
Pluralis Research is seeking a Research Engineer to build and own the inference stack for their innovative Protocol Learning system, enabling decentralized AI model training and serving on consumer devices. This role involves designing novel algorithms for fast and reliable inference over the public internet and ensuring the pipeline’s performance for both reinforcement learning training and model serving. The ideal candidate will have experience shipping serving systems and a passion for decentralized and trustless AI.
What you'll do
- Own the inference stack end-to-end.
- Invent algorithms for fast inference on consumer hardware.
- Maintain a fast and reliable rollout pipeline for RL training.
- Transition the pipeline to a serving layer for trained models.
What they're looking for
- Serving systems
- Distributed inference
- LLM serving
- Pipeline parallelism
- Low-bandwidth networking
- Algorithm design
- Reinforcement Learning (RL)
- MLX
Benefits
- Equity-heavy package
- Remote-first culture
- Visa sponsorship (US or Australia)
- Flexible work environment
Opens the application — the Jobs AI extension fills it for you. Set up autofill
Opens the official application on the employer’s site. No login required.
Pluralis Research
Likely interview questions
- Describe a time you shipped a serving engine or large-scale inference system. What were the biggest challenges and how did you overcome them?
- Explain your experience with low-bandwidth, high-latency networks. What strategies did you use to optimize performance?