micro1, a US company that builds training data and evaluations for AI labs, lists two full-time, remote infrastructure roles on its own team.
Member of Technical Staff, Model Training (infra and platform). Build shared infrastructure for supervised fine-tuning, RL and on-policy distillation across language, multimodal and robotics workloads; design distributed pipelines that coordinate GPU workers, rollouts, teacher inference, scoring and simulation; optimize GPU utilization, memory and storage. The listing asks for distributed systems depth, Python plus a systems language (C++, Rust or Go), and hands-on GPU training or inference infrastructure. It names PyTorch Distributed, DeepSpeed, Ray, vLLM, CUDA and Kubernetes as preferred. Listed salary $200,000 to $320,000 a year.
Member of Technical Staff, Infrastructure. Design secure, resilient cloud infrastructure for AI products; build AI-driven workflows for provisioning, deployment, monitoring and incident response; own logging, metrics, tracing and alerting. The listing asks for staff-level experience in infrastructure, SRE or platform engineering on AWS, GCP or Azure. Listed salary $220,000 to $300,000 a year.
Applications go through micro1’s own site.
Disclosure: I run AI Trainer Guide, an independent guide to AI training work. I am not micro1 or its recruiter. The pages above link to micro1’s listings with a referral link, so micro1 may pay us if it hires you, at no cost to you. Questions about the roles themselves are best put to micro1.
Cisco, AI Trainer Guide