AI Training Systems Engineer - Distributed GPU Infra Remote

Tilted is building an operating system for an agent economy. As a Systems Engineer, Training Infrastructure, you will stand up the training stack and scale a multi-node GPU cluster, owning MFU metrics and fault-tolerance for weeks-long runs. You’ll work closely with the Founding Pretraining Lead and help shape the architecture of TiltedOS and the mesh-based job system. This is not a cog role: you’ll touch from kernel to scheduler, building a frontier training program that runs as typed jobs on #J-18808-Ljbffr