Research Engineer - Geo-Distributed Inference
Core
Build and own the end-to-end inference stack for a decentralized, permissionless RL training and serving pipeline running on consumer-grade hardware over the public internet.
Role type
Senior IC Research Engineer (Distributed Systems & Inference)
Builds
High-throughput inference pipeline for RL training rollouts and future model serving
Domain
Decentralized AI, Distributed Systems, Reinforcement Learning
Deliverable
production ML models
Required skills
Distributed inference systems, Pipeline parallelism, Low-bandwidth networking, Failure handling, System design, Algorithm design, Production deployment, P2P networking, NAT traversal
Preferred skills
RL post-training, Apple silicon/MLX, Open-weight AI labs experience
Technologies
Macs, Consumer GPUs, Public Internet, MLX
Responsibilities
Design and validate new algorithms for fast inference on consumer hardware over the public internet, Build and maintain the inference stack including placement, routing, and transport, Ensure reliability of the rollout pipeline for RL training and future model serving
Seniority
Senior, hands-on IC
