Inference Infrastructure Architect (Remote)

At Telnyx, the team is focused on building and managing a global, multi-cloud IP network that supports seamless connectivity for various applications and devices. The company emphasizes innovation in connectivity solutions, automation, and modernization of infrastructure, fostering a profitable environment that encourages continuous learning and growth. The organization aims to shape the future of borderless communication and interconnected technology.

The role of the Inference Infrastructure Architect involves overseeing the company's hardware fleet used for AI inference, optimizing throughput, and reducing costs while maintaining performance standards. The person will work on deploying serverless inference solutions, managing Kubernetes on bare metal, and integrating open-source components for large language model serving. Daily tasks include system tuning, benchmarking, and ensuring reliable, cost-effective AI inference operations.

Candidates should have experience with production-level large language model serving, Kubernetes on GPU clusters, and performance engineering. Skills in Python, Go, Linux, networking, and storage are essential, along with the ability to work with community projects and documentation. The role is remote, with a focus on building scalable, efficient inference infrastructure that supports enterprise and open-weight AI models.