SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Relace is building a globally distributed inference cloud that processes 800B+ tokens per day. The company competes on price, speed, and reliability against much larger competitors, with customers and revenue today. Relace is venture-backed (Series A $23M led by Andreessen Horowitz), growing 3x in the last three months, and regularly ranks as a top provider by volume on OpenRouter.
In this role, you will work on the core distributed systems that power Relace's globally distributed inference platform. Specifically, you'll focus on the routing layer that decides which machine handles each request, capacity management when demand exceeds available resources, and fleet orchestration across multiple clouds as demand fluctuates. The technical stack centers on Rust services and Kubernetes for managing a multi-cloud fleet.
You'll own critical infrastructure components where performance directly impacts the business: a 10% improvement in system speed translates to roughly 10% more revenue. This is a senior individual contributor role with significant responsibility and room to grow within a small, high-impact team.
Key responsibilities include designing and maintaining the request routing system, implementing queue and load-shedding logic, managing multi-cloud fleet scaling, and ensuring reliability under high load. You'll work closely with the infrastructure and product teams to optimize latency, throughput, and cost.
REQUIREMENTS:
- 5+ years building production systems where latency and failure behavior were critical (serving systems, control planes, schedulers, databases, networking, trading systems). Data pipelines and analytics do not count.
- Deep proficiency in a systems language: Rust, C++, or Go.
- Demonstrated ownership of production systems running under load, with the ability to discuss a production failure you debugged in detail.
- Kubernetes expertise at the level of having extended the platform, particularly on low-reliability infrastructure.
NICE TO HAVE:
- Experience building or significantly contributing to a scheduler or load balancer as a core project.
- Work with spot or preemptible compute capacity.
- Any exposure to AI serving systems (domain knowledge is optional; distributed systems instincts are not).
