The Microsoft Azure cloud is a global platform that empowers every person and organization on the planet to achieve more. We host services at the scale of Bing, Teams, SharePoint, and SQL Azure. Customers use the power of Azure to host millions of multiplayer gaming sessions and solve large computational problems. Workloads range from \"born-in-the-cloud\" services to enterprise applications migrated from customer datacenters.

The One Fleet Control Plane, the kernel of the Azure cloud, is a fault-tolerant distributed system built on commodity datacenter hardware. Within the One Fleet Control Plane, the Compute Allocator team develops services to assign customer workloads to physical compute servers. As Azure expands, efficient utilization of our resources becomes a vital part of the business. Allocation services must maintain a high throughput of workload assignments to meet customer elasticity requirements but also pack as efficiently as possible. Most importantly, allocation decisions must be correct, respecting placement constraints to meet customer performance and availability requirements. The Compute Allocator team develops the Protean allocation technology described at https://www.microsoft.com/en-us/research/publication/protean-vm-allocation-service-at-scale/.  

In this role, you will enhance the Azure Compute allocation services as they are applied to new customer scenarios, collaborating with partner teams across Azure. You will tune packing algorithms to improve customer experience and platform efficiency, and you will drive improvements to availability, scalability, and operability. If you are interested in distributed systems, if you enjoy algorithms and efficiency, if you seek opportunities for high impact on Microsoft's core business, then this is the role for you. Team is comfortable having remote candidate within Unites States.