We're looking for a hands-on Operations Manager to own the operational and analytical supply side of our GPU fleet. Key focus areas: GPU fleet lifecycle, health, observability, utilization monitoring, and remediation across our neocloud and bare metal environments.

We contract for a fixed amount of compute capacity. GPUs drift from healthy to unhealthy over time, and this role minimizes that downtime to keep the maximum number of GPUs healthy at any given moment.

This is an operator role, not people management. You'll drive execution through clear processes, metrics, reporting, vendor coordination, and cross-functional alignment..

Core Responsibilities:

Drive suppliers to keep the maximum amount of the GPU fleet online and healthy.

Maintain a live reconciliation of contracted vs. provisioned vs. healthy vs. utilized capacity, broken out by supplier and by cluster maximizing the number of healthy GPUs.

Supplier-attributed fleet health accountability: own replacement SLAs, mean time to repair (MTTR), and RMA cycle times for every in-scope supplier.

SLA monitoring, credit claims, and remedy enforcement: track SLA performance against contract terms, file and pursue credit claims, and drive remediation plans when suppliers fall short.

Drive internal communications where suppliers need to perform maintenance to ensure all Baseten stakeholders are aware of activities that impact availability.

Scope and Approach

Flexibility: this list covers the core of the role, not the limit of it. You'll be asked to take on adjacent work as the function evolves and as new gaps surface.

Ownership mindset: we need someone who treats "whatever it takes" as a genuine operating principle, not a line in a job posting. If something falls outside a defined lane but inside the overall goal of closing the capacity gap, it's yours to pick up.

Capacity Operations Manager

Responsibilities

Qualification

Required

Preferred

Benefits

Company

Baseten

More roles at Baseten