Software Engineer, Training Infrastructure
Core
Build distributed training infrastructure for frontier AI models and the Tinker fine-tuning API, managing scheduling, fault tolerance, and optimization across large GPU clusters.
Role type
Senior IC distributed systems engineer (ML training infrastructure)
Builds
Distributed training stack for supervised and reinforcement learning jobs, Tinker API backend
Domain
AI/ML infrastructure, large-scale GPU clusters
Deliverable
production ML models
Required skills
Python, C++/Rust/Go, distributed systems, PyTorch, performance profiling, memory optimization, fault tolerance, multi-tenant isolation
Preferred skills
LLM training at scale, mixture-of-experts, LoRA, Kubernetes/Slurm, observability tooling, open-source ML contributions
Technologies
PyTorch, Kubernetes, Slurm
Responsibilities
Design and operate distributed training infrastructure; optimize throughput and cost via profiling and tuning; support new model architectures; build reliability features like failure detection and recovery; enable multi-tenant fine-tuning; collaborate with API and product teams
Seniority
Senior, hands-on IC